Almost Disjunct Codes in Large Scale Multihop Wireless Network Media Access Control

Size: px
Start display at page:

Download "Almost Disjunct Codes in Large Scale Multihop Wireless Network Media Access Control"

Transcription

1 Almost Disjunct Coes in Large Scale Multihop Wireless Network Meia Access Control D. Charles Engelhart Anan Sivasubramaniam Penn. State University University Park PA 682 Abstract We iscuss the application of coing constructs to the problem of raio transmission scheuling in a hoc wireless networks. We are particularly intereste in the ability to scale such a hoc wireless networks to thousans of noes. Current meia access protocols o not scale well to larger networks than a few ozen noes. We examine the possible ways to apply superimpose coes an almost isjunct coes to these networks meia access layer. We analyze how well one can buil coes that are not as powerful as superimpose coes but have better parameters for the case of a hoc networks. We look at the local network performance that a particular noe can expect using these coes. We also examine the system wie performance using an approximate simulation of thousans of a hoc wireless noes. Introuction We woul like to improve the performance of meia access control (MAC) in wireless a hoc networks. We are intereste in performance both in terms of quality of service measures like throughput an latency but also in terms of measures that will affect battery life, such as collision rate. In orer to transmit ata a wireless noe nees to be sure that no other noes in its raio range are transmitting at the same time so the transmission can be hear by its neighbors. Traitionally MAC protocols can be classifie as either centralize or istribute algorithms. Centralize protocols like TDMA suffer from low spatial utilization as generally only one noe in the system is allowe to transmit at a time, but guarantee no conflicts. Distribute algorithms offer the possibility of spatial reuse but suffer from significant contention problems even in moerately loae systems, an usually cannot guarantee reliable elivery. We are particularly intereste in how these systems will scale to thousans of noes or more. Since these noes are usually mobile evices an thus are resource constraine, especially with regar to battery life, we will be more intereste in the efficiency of the system than throughput or latency. Recently avance protocols base on the ieas of TDMA, reliable elivery an little contention, have starte to appear. These new protocols use avance coing structures to reuse the space an still guarantee reliable elivery. These systems were first iscusse in [2] an refine in [3]. Both protocols use constructions of orthogonal coes to ensure that a conflict free scheule can be obtaine without knowlege of the unerlying network topology. Both stuies analyze the theoretical throughput of these systems an the latter more closely analyzes guarantee throughput as well as scheuling elays. A recent paper by Syrotiuk, et al. [6] generalizes the use of orthogonal arrays for these systems an analyzes both the probability of successful transmission an throughput of such systems built from general orthogonal arrays of strength 2 an 3. The remainer of this paper is organize as follows. Section 2 goes over some basic coing constructs. The ways they can be applie to wireless networks is iscusse in Section 3. Section 4 examines the localize network performance when using superimpose coes. An approximate simulation of a large scale wireless network is use to compare an almost isjunct coe to an a hybri protocol in Section 5. 2 Orthogonal arrays an Superimpose coes In general we are going to use a binary matrix to represent each noes transmission scheule. The rows of the ma-

2 trix correspon to the given time-slots that noes can transmit in. Each column of the matrix represents a single noe. Thus each time a appears in a row i column j then noe j is allowe to transmit uring time-slot i. The scheule is then repeate inefinitely. The number of columns of such a matrix will limit the number of noes in a system. In general we will let n be the number of noes in the system. 2. Orthogonal arrays Previous work [6] has use orthogonal arrays to construct these scheules. Def. OA(t, k, v) is a k v t matrix using the symbols {,, 2...v } is a strength t orthogonal array if for any set of t of its rows each t-tuple of column elements appears exactly once, for that row set. One important property of such an orthogonal array is that each column, or coewor, of OA(t, k, v) intersects every other column in fewer than t positions. Using this property we can buil a transmission scheule. First no two columns intersect in more than t positions, so given a set of D other columns the given column iffers from them in at least k D(t ) positions. As long as this ifference is positive we can say that there is at least one row where the given column is ifferent from all of any set of D other columns. If we assume this D is boun for the number of neighbors any given noe can have we can then assure each noe has a time-slot in which it is allowe to talk when none of its neighbors are. The symbols {,, 2...v } are mappe to the set of ientity vectors. Using this binary vectorization we can see that as long as 2 symbols are ifferent then uring that row the 2 noes will not interfere with each other. So using OA(t, k, v) we can hanle up to v t noes an a scheule inclues k v time-slots. 2.2 Superimpose coes We can also construct viable scheules using superimpose coes [4]. We say one binary vector a covers another b iff for every in b there is a in the same position in a. Def. A binary matrix is -isjunct iff it has the property that the union of any coewors oes not cover any coewor not in that set of coewors. We can then construct a scheule using such a -isjunct superimpose coe, because we know there is at least one row where there exists a corresponing to a given noe an there are all s for all of its neighbors. Each noe is guarantee a time-slot where it can talk an none of its neighbors may. We can then see that what we constructe with the binary vectorize orthogonal arrays was a D-isjunct superimpose coe. 2.3 Almost isjunct coes We woul like to relax the property of having an absolute guarantee that there exists a time-slot in which every noe is allowe to talk with no interference. To o so we wish to consier α-almost -isjunct coes []. Def. A binary matrix is α-almost -isjunct if for any ranomlyselecte set of columns, the probability that they cover no other column is at least α. So we can say that a -almost-isjunct matrix is exactly a -isjunct superimpose coe. These matrices have the property that each noe has probability α that it has a time-slot where it can have interference free communication The Matrix * Operator One way of increasing the number of noes that a coe can hanle without increasing its scheule length much is by using the matrix * operator. The * operator applie to a single matrix prouces another matrix that has every orere pair of columns from the first matrix. Given a matrix, M, the matrix M then has twice as many rows, but the number of columns is square. It is worth noting that given any matrix M we cannot guarantee that M is even 2-isjunct, but it may very well be highly α-almost -isjunct for a given. 3 Superimpose coes in wireless networks The scheules in previous work have been built on OA(2,v,v) in [2] an enhance to OA(2,v +,v) in [6] for an increase in throughput. The resulting scheules will then have v 2 coewors of length either v 2 or v 2 + v an can hanle up to v neighbors, that is it will be a v-isjunct superimpose coe. In [6] they also iscuss the aition of strength 3 orthogonal coes an show that for small numbers of neighbors that they can significantly increase throughput. They again focus on orthogonal arrays where k = v +, in oing so they increase the number of coewors to v 3.However,by not increasing the length of each coewor they can have a maximum of (v +)/2 neighbors, which roughly cuts in half the power of the superimpose coe. In orer to not reuce the power of the resulting superimpose coe, an thus the maximum number of neighbors allowe in the system they woul have to fin an orthogonal array with k approximately twice v. This woul yiel a scheule with coewors of length approximately 4v 2. It can be seen that in the initial constructions using orthogonal arrays that the length of each coewor an the number of coewors is asymptotically the same. This means that a scheule must be as long as the number of 2

3 noes in the system. So in regar to scheule length these scheules are no better than traitional TDMA. They o, however, significantly increase the amount of ata that it is possible to transmit uring that time perio. If we plan to use these scheules for very large systems, they are also going to be require to have a very large scheule length, approximately the size of the number of noes. For example in a system with one million noes we woul nee to construct OA(2,, ) (is this easy to o?). We coul then hanle neighbors among any of the one million noes, this may seem a bit too powerful for what our nees may be, perhaps we woul nee to only be able to hanle neighbors. More importantly each scheule has one million time-slots. It may be the case that certain noes are only allowe to transmit uring one small section of that scheule, because they are only guarantee one time-slot in which none of their neighbors are allowe to transmit. This woul possibly lea to large latencies, an if we want to increase the number of noes further it coul lea to near starvation. 3. Incience matrix construction of superimpose coes One well known [5] construction of superimpose coes is base on the iea of incience matrices of sets. Each row an column in such a matrix represents a given subset of m objects. We then say there is a in position (i.j) iff the set corresponing to row i is completely containe in the set corresponing to column j. A-isjunct superimpose coe can be constructe with ( ) ( m rows (time-slots) an m ) s columns (noes) where m<s<. Consequently we can have superimpose coes that asymptotically grow much faster in the number of coewors than the scheule length. We can see that even with a relatively small these coes are only going to be suite for large systems. The other main factor that will contribute to their performance will be the ensity of the coes, that is the percentage of ones that appear in the coe, that is the percentage of time the noes are allowe to transmit. 3.2 Constructing Almost isjunct coes Our initial metho for constructing almost isjunct coes is to simply remove some percentage of the rows. By oing this we are cutting the scheule length but we are also potentially reucing the power of the coe. We nee to be careful that we o not remove rows in a manner which overly egraes the power of the coe, or one that penalizes some noes significantly. We will remove rows uniformly at ranom in orer to try an minimize these cases. This is not guarantee to minimize conflicts, but on average shoul penalize no coewor(noe) more than any other. With this reuction in scheule length we are not going to be able to make the same guarantees as before but hopefully these scheuling conflicts can be minimize to the point that they only occur with very low probability. The way in which the superimpose were initially constructe will make the ensity of the coe remain constant when removing rows. alpha m=2, s=6, =3 Incience Matrix Coe Frame Length Figure. The α- almost isjunct property as rows are remove from the superimpose coe. Figure shows how alpha changes as we remove rows in the superimpose coes constructe with the incience matrix construction. There were 22 rows in the initial coe, notice that with only /4 of the original rows that the coe is still.5-almost 3 isjunct, an with half the rows remove it is still about.85-almost 3 isjunct. We can also use a less powerful superimpose coe than the number of neighbors. For example with a power 3 superimpose coe we can only guarantee success in the case of 3 or less neighbors, but it is very likely that we will still have some success with more than 3 neighbors. Figure 2 shows the alpha values in the case of a =3 superimpose coe with up to 3 neighbors. Using a coe that is only guarantee to be 3 isjunct it is.95-almost 5 isjunct, an.6-almost 3 isjunct. It appears there are great gains to be ha if we are willing to eal with a chance that our coe oes not strictly meet the isjunct property of a superimpose coe. In a wireless network this woul mean that on a given scheule there is a α chance that a noe woul not be guarantee a time slot where only it was talking amongst its neighbors. 3.3 Parameter Comparison Figure 3 compares the scheule length of 7-isjunct superimpose coes constructe with orthogonal arrays an the incience matrix construction, using completely isjunct coes. The OA(2,v,v) coe has the longest scheule 3

4 alpha m=2, s=6, =3 Incience Matrix noes scheule length neighbors TDMA n n n orthogonal coes v t k v k t OA(2,v,v) coes v 2 ) v ( 2 v m ) incience matrix almost isjunct ( m s ( m s ) β ( ) m?? Table. Parameters of various MAC protocol coes Desire Figure 2. The α- almost isjunct property with a lower power superimpose coe. lengths of the three over all values of network sizes. The OA(2, 8,v) orthogonal coe performs better at lower numbers of noes, but when the network size is increase significantly the growth rate of the incience matrix superimpose coe has shorter scheule lengths. Table 3.3 shows the various parameter constraints for the coes iscusse. Scheule Length OA(2,8,v) Incience matrix 7 isjunct TDMA an OA(2,v,v) Noes the coe. So in orer to be able to talk about how it shoul perform we will want expressions for the ensity (the fraction of ones) of each coe. The simplest way to look at the ensity of a superimpose coe constructe via an incience matrix is not to look at the ones, but instea the situations that will lea to zeros in the coe. It is clear that a superimpose coe constructe via the incience matrix metho will have the same number of ones or zeros in each column, an similarly in each row. So to fin the ensity of the coe we only nee to look at one row or column. Recall that a row in such a superimpose coe correspons to a set of objects chosen from m objects, similarly columns correspon to a set of s objects chosen from those same m objects. Looking at any column we can see that the only time a one will appear in any position is when the set corresponing to the row is completely containe in the set corresponing to that column. There are exactly ( s ) sets of size that are completely containe in the s-set corresponing to a given column. Since there are exactly elements in a given column the coe ensity is then ) (s. This ensity ( m ) remains constant across rows of the matrix as well as the columns. Figure 3. The scheule length of various 7- isjunct superimpose coes. =3 =5 =7 4 Localize Network Performance Coe Density 2 In orer to unerstan how a wireless network will perform we will first look at the theoretical network performance in the neighborhoo aroun a given noe. 4. Coe Density For superimpose coes one of the primary characteristics that will effect network performance is the ensity of Noes (m=[2,5], s=m/2) Figure 4. Coe ensity as the number of noes increases. 4

5 Figure 4 shows how the ensity of the coe changes as the number of noes (coewors) in the system increases for 3, 5 an 7-isjunct superimpose coes, constructe with the incience matrix construction. The values for m are in the interval [2,5], an the s values are always chosen to be half of a given m, remember that m an s etermine the maximum number of noes for a given superimpose coe. Notice that as the power of the superimpose is increase the ensity of the coe is lower. Also as the number of noes increases the ensity increases slightly. A successful transmission then occurs when the noe in question is allowe to transmit an all of its neighbors are silent. Thus the probability of successful reception for a noe that wishes to communicate is simply P r = P a ( P a ) P b. Figure 6 shows how the expecte reception ratio changes as the ensity of the coe increases, with a fixe number of neighbors. Notice that when all noes wish to communicate all the time the esire coe ensity is higher than in a more lightly loae system..5 m=4, s=2 5.8 % transmission probability 2% transmission probability.6.35 Coe Density.3 5 Reception Probability Coe Density Figure 5. Coe ensity as the power,, ofthe superimpose coe increases. Figure 6. The reception probability as the ensity of the coe changes. Figure 5 shows how the ensity changes with the power,, of the superimpose coe. As we saw before as the power of the coe increases its ensity ecreases. This ensity remains constant as we remove rows to from an almost isjunct coe. The ensity for stanar TDMA is clearly /n an for orthogonal arrays is /v which in the case of OA(2,v,v) coes is / n. These coes will all be somewhat inflexible if we wish to change the ensity of the coe to suit a particular network configuration, but we have some choice in coe ensity when using an incience matrix superimpose coe. 4.2 Throughput Throughput at the MAC layer is a simple function of the reception success rate for a given noe that wishes to transmit. We erive a simplifie moel of reception success that ignores complex propagation etails. Let P r be the probability of a successfully receive transmission. P r will epen on several factors. The probability that the noe is allowe to transmit uring that times lot, P a, is the first factor. Secon, this success rate will also epen on the number of neighbors,, that noe has as well as the likelihoo that a given noe will wish to transmit, P b. Figure 7 shows how the reception probability changes with varying the number of neighbors, for various coe ensities. As expecte increasing the number of neighbors reuces the probability of a successful communication, but notice how the lower ensity coes become more esirable as the number of noes increases. Reception Probability ensity=.5 ensity= ensity=.3 ensity= ensity= Figure 7. The reception probability as more neighbors are ae to the system. Figure 8 shows how varying both the number of neigh- 5

6 bors an transmission probability effects the reception probability. Reception Probability Transmission Probability Figure 8. The reception probability varying with both the number of neighbors an their transmission probability. 5 Simulation Although looking at the performance localize aroun a group of noes may give us some insight into how to optimize such networks we must look at the whole network working together in orer fully analyze any MAC protocol. To o this we use an approximate simulation that moels a complete, large scale, wireless network. We use Matlab [7] to simulate these networks. The noes in the networks we simulate are all static, that is they o not move over the course of the simulation. They are istribute uniformly at ranom over a square area. Raio propagation is hanle by simply calculating the istance between pairs of noes an if they are within some threshol istance of each other they can communicate, an otherwise they cannot. Messages are generate using intervals etermine by using an exponential istribution, governe by a parameter λ. Asλ increases the time between message generation shortens, thus there is more loa in the system. Messages are sent from any ranom noe to another ranom noe. Routes for the messages are mae using a geographic algorithm for simplicity, the next hop of a message is sent to the noe in its communication range that is geographically closest to its final estination. The communication ranges were set so that there woul be 6 neighbors per noe on average. We use a mbps raio moel, using 52 byte packets. We ignore all packets elivere uring an initial start up winow, an most of the simulations ran for a few minutes of simulate time. We ran two experiments for each raio moel. First we varie the size of the network an kept the loa (etermine by the parameter λ in the exponential traffic generator) constant. Secon we kept the network size constant an varie the loa in the system generate by the exponential traffic generator. For each experiment we measure the total number of routing layer packets elivere, the collision rate an the average latency for routing layer packets. Note that the collision rate is going to have a irect effect on the power use by noes in the system, as collisions are simply transmissions where power was consume an no work was one. 5. MAC Protocols We compare three ifferent MAC protocols for each of the simulations. First we use a protocol. We use time intervals that were smaller than the time to transmit a packet (/25th the time to sen a packet) for raio contention. In this moel a given transceiver only tries to communicate if no one is currently using the channel. If there is a collision, or anytime that the raio wants to talk after it has hear another transceiver talking it must wait a ranom number of the smaller time intervals before it can try to transmit. That ranom amount of timer intervals is base on an exponential back off timer. The back off winow has a minimum size of 8 time intervals an a maximum of 256, with it oubling every time the noe cannot communicate, an resetting to 8 every time it communicates successfully. We also use a coe base MAC protocol. For simplicity sake we use a very simple coe that has goo α -almost isjunct properties. We use the minimum In matrix that coul support the number of noes use in the simulation, for example if there were simulate noes we woul have use I, which has a scheule length of 2 an a coe ensity of.. We i not use any smaller time intervals for this simulation, an the noes ability to communicate uring a transmission winow is completely base on the column of the coe corresponing to that noe. In orer to try to reuce the latency problems associate with the coe base MAC protocol we use a hybri approach that use the same coe use for the coe base protocol, but also incorporates the smaller time intervals. So a given noe can start transmitting uring a time slot where it has a an continues transmitting until it elivers the whole packet (which takes 25 of the smaller time units). There is also a carrier sense mechanism use so that no noe will try to start communicating when another noe is alreay transmitting. 6

7 25 9 coe base hybri coe base hybri Messages passe Average Latency Noes Noes Figure 9. Routing layer messages passe as the size of the network grows. Figure. Average latency of routing layer packets as the size of the network grows. 5.2 Network size scaling First we examine how each of the protocols scales as the network grows in size. The parameter for the traffic generator, λ, was kept constant at.. Figure 9 shows the number of messages passe for each protocol. The coe base protocol starts out the worst of the 3 when there are fewer noes in the system, but it overtook the approach aroun 5 noes. It is worth noting the hybri approach completely faile (near throughput, tocols perform almost ientically until about 3 noes are reache, where the coe base protocol starts to perform significantly better than. Overall the coe base protocol seems to outperform the base an hybri protocols as the size of the network grows. 5.3 Network traffic scaling We fixe the number of noes in the system at an varie the rate at which traffic was generate by varying λ..5 5 coe base hybri 4 coe base hybri 2.35 Collision Rate.3 5 Messages passe Noes λ 6 8 x Figure. The collision rate as the size of the network grows. Figure 2. The number of routing layer messages passe as more traffic is generate. Figure shows the collision rate for each protocol. Notice that both the coe base an protocols have a significantly lower collision rate than the hybri system. The coe base protocol has a lower collision probability over the range of network sizes. Figure shows the average latency of routing level packets for each protocol. The coe base an pro- Figure 2 shows the number of messages passe for each protocol as network traffic increases. As we woul expect as more traffic is generate more messages are passe overall. The hybri protocol seems to pass the most messages with the an coe base protocols performing similarly. Figure 3 shows collision rate for each protocol as network traffic increases. The hybri protocol reaches satura- 7

8 .9 coe base hybri 6 Summary an Conclusions Collision Rate λ 6 8 x Figure 3. The collision rate as more traffic is ae to the system. tion with the lowest traffic rate of the three protocols. The protocol also reaches saturation soon after, but the coe base system seems to taper off with a collision rate aroun 3% coe base hybri We examine several coing constructs for use in multihop wireless networks. Of the currently use coes the incience matrix constructe superimpose coes ha better growth rates when scaling to large numbers of noes, an offere some flexibility in terms of coe ensity. There appear to be great gains possible if we are willing to relax the isjunct property of superimpose coes an use almost isjunct coes. We also examine how various properties of the coes use woul effect localize network performance. As we increase the number of noes in the system a very low ensity coe appears to be esirable. Increasing the reception probability locally woul also have a two-fol effect, it not only woul increase the network performance in terms of latency an throughput, but will also reuce the power costs of the system by avoiing collisions. In our simulations there was not a clear winner in terms of throughput or latency, but in terms of collision rate it appeare the almost isjunct coe performe the best. The hybri approach showe some potential, but also showe its limitations when scaling the number of noes in the system. References Average Latency λ 6 8 x Figure 4. The average routing layer packet latencyasmoretrafficisgenerate. Figure 4 shows average routing layer packet latency for each protocol as network traffic increases. As expecte, in general when we pump more traffic into the system the average latency increases. The hybri protocol seems to perform best of the 3 protocols with the an coe base protocols performing somewhat similarly, until the protocol is near saturation. Although the other protocols may seem to have favorable throughput numbers it seems the coe base system is better because it oes not saturate as quickly as the network traffic is increase. Throughout the situations teste as network characteristics scale up the coe base system, using approximate superimpose coes, outperforms the other two methos. [] A.J.MACULA, V.V.RYKOV, S. Y. Trivial two-stage group testing for complexes using almost isjunct matrices. Discrete an Applie Mathematics 37 (24), [2] CHLAMTAC, I.,AND FARAGÓ, A. Makingtransmission scheules immune to topology changes in multihop packet raio networks. IEEE/ACM Transactions on Networking 2, (February 994), [3] JU, J.-H., AND LI, V. O. K. An optimal topologytransparent scheuling metho in multihop packet raio networks. IEEE/ACM Transactions on Networking 6,3 (June 998), [4] KAUTZ, W.H.,AND SINGLETON, R. Nonranom binary superimpose coes. IEEE Transaction on Information Theory (October 964), [5] MACULA, A. J. Simple construction of -isjunct matrices with certain constant weights. Discrete Mathematics 62 (996), [6] SYROTIUK, V.R.,COLBOURN, C.J.,AND LING, A. C. H. Topology-transparent scheuling for manets using orthogonal arrays. Preprint, May 23. [7] THE MATHWORKS, I. Matlab. 8

Improving Spatial Reuse of IEEE Based Ad Hoc Networks

Improving Spatial Reuse of IEEE Based Ad Hoc Networks mproving Spatial Reuse of EEE 82.11 Base A Hoc Networks Fengji Ye, Su Yi an Biplab Sikar ECSE Department, Rensselaer Polytechnic nstitute Troy, NY 1218 Abstract n this paper, we evaluate an suggest methos

More information

Questions? Post on piazza, or Radhika (radhika at eecs.berkeley) or Sameer (sa at berkeley)!

Questions? Post on piazza, or  Radhika (radhika at eecs.berkeley) or Sameer (sa at berkeley)! EE122 Fall 2013 HW3 Instructions Recor your answers in a file calle hw3.pf. Make sure to write your name an SID at the top of your assignment. For each problem, clearly inicate your final answer, bol an

More information

On the Placement of Internet Taps in Wireless Neighborhood Networks

On the Placement of Internet Taps in Wireless Neighborhood Networks 1 On the Placement of Internet Taps in Wireless Neighborhoo Networks Lili Qiu, Ranveer Chanra, Kamal Jain, Mohamma Mahian Abstract Recently there has emerge a novel application of wireless technology that

More information

Lab work #8. Congestion control

Lab work #8. Congestion control TEORÍA DE REDES DE TELECOMUNICACIONES Grao en Ingeniería Telemática Grao en Ingeniería en Sistemas e Telecomunicación Curso 2015-2016 Lab work #8. Congestion control (1 session) Author: Pablo Pavón Mariño

More information

An Adaptive Routing Algorithm for Communication Networks using Back Pressure Technique

An Adaptive Routing Algorithm for Communication Networks using Back Pressure Technique International OPEN ACCESS Journal Of Moern Engineering Research (IJMER) An Aaptive Routing Algorithm for Communication Networks using Back Pressure Technique Khasimpeera Mohamme 1, K. Kalpana 2 1 M. Tech

More information

Generalized Edge Coloring for Channel Assignment in Wireless Networks

Generalized Edge Coloring for Channel Assignment in Wireless Networks Generalize Ege Coloring for Channel Assignment in Wireless Networks Chun-Chen Hsu Institute of Information Science Acaemia Sinica Taipei, Taiwan Da-wei Wang Jan-Jan Wu Institute of Information Science

More information

Intensive Hypercube Communication: Prearranged Communication in Link-Bound Machines 1 2

Intensive Hypercube Communication: Prearranged Communication in Link-Bound Machines 1 2 This paper appears in J. of Parallel an Distribute Computing 10 (1990), pp. 167 181. Intensive Hypercube Communication: Prearrange Communication in Link-Boun Machines 1 2 Quentin F. Stout an Bruce Wagar

More information

Classifying Facial Expression with Radial Basis Function Networks, using Gradient Descent and K-means

Classifying Facial Expression with Radial Basis Function Networks, using Gradient Descent and K-means Classifying Facial Expression with Raial Basis Function Networks, using Graient Descent an K-means Neil Allrin Department of Computer Science University of California, San Diego La Jolla, CA 9237 nallrin@cs.ucs.eu

More information

Shift-map Image Registration

Shift-map Image Registration Shift-map Image Registration Svärm, Linus; Stranmark, Petter Unpublishe: 2010-01-01 Link to publication Citation for publishe version (APA): Svärm, L., & Stranmark, P. (2010). Shift-map Image Registration.

More information

Queueing Model and Optimization of Packet Dropping in Real-Time Wireless Sensor Networks

Queueing Model and Optimization of Packet Dropping in Real-Time Wireless Sensor Networks Queueing Moel an Optimization of Packet Dropping in Real-Time Wireless Sensor Networks Marc Aoun, Antonios Argyriou, Philips Research, Einhoven, 66AE, The Netherlans Department of Computer an Communication

More information

Message Transport With The User Datagram Protocol

Message Transport With The User Datagram Protocol Message Transport With The User Datagram Protocol User Datagram Protocol (UDP) Use During startup For VoIP an some vieo applications Accounts for less than 10% of Internet traffic Blocke by some ISPs Computer

More information

Shift-map Image Registration

Shift-map Image Registration Shift-map Image Registration Linus Svärm Petter Stranmark Centre for Mathematical Sciences, Lun University {linus,petter}@maths.lth.se Abstract Shift-map image processing is a new framework base on energy

More information

Probabilistic Medium Access Control for. Full-Duplex Networks with Half-Duplex Clients

Probabilistic Medium Access Control for. Full-Duplex Networks with Half-Duplex Clients Probabilistic Meium Access Control for 1 Full-Duplex Networks with Half-Duplex Clients arxiv:1608.08729v1 [cs.ni] 31 Aug 2016 Shih-Ying Chen, Ting-Feng Huang, Kate Ching-Ju Lin, Member, IEEE, Y.-W. Peter

More information

Architecture Design of Mobile Access Coordinated Wireless Sensor Networks

Architecture Design of Mobile Access Coordinated Wireless Sensor Networks Architecture Design of Mobile Access Coorinate Wireless Sensor Networks Mai Abelhakim 1 Leonar E. Lightfoot Jian Ren 1 Tongtong Li 1 1 Department of Electrical & Computer Engineering, Michigan State University,

More information

Computer Organization

Computer Organization Computer Organization Douglas Comer Computer Science Department Purue University 250 N. University Street West Lafayette, IN 47907-2066 http://www.cs.purue.eu/people/comer Copyright 2006. All rights reserve.

More information

Coupling the User Interfaces of a Multiuser Program

Coupling the User Interfaces of a Multiuser Program Coupling the User Interfaces of a Multiuser Program PRASUN DEWAN University of North Carolina at Chapel Hill RAJIV CHOUDHARY Intel Corporation We have evelope a new moel for coupling the user-interfaces

More information

Overview : Computer Networking. IEEE MAC Protocol: CSMA/CA Internet mobility TCP over noisy links

Overview : Computer Networking. IEEE MAC Protocol: CSMA/CA Internet mobility TCP over noisy links Overview 15-441 15-441: Computer Networking 15-641 Lecture 24: Wireless Eric Anerson Fall 2014 www.cs.cmu.eu/~prs/15-441-f14 Internet mobility TCP over noisy links Link layer challenges an WiFi Cellular

More information

Here are a couple of warnings to my students who may be here to get a copy of what happened on a day that you missed.

Here are a couple of warnings to my students who may be here to get a copy of what happened on a day that you missed. Preface Here are my online notes for my Calculus I course that I teach here at Lamar University. Despite the fact that these are my class notes, they shoul be accessible to anyone wanting to learn Calculus

More information

Lecture 1 September 4, 2013

Lecture 1 September 4, 2013 CS 84r: Incentives an Information in Networks Fall 013 Prof. Yaron Singer Lecture 1 September 4, 013 Scribe: Bo Waggoner 1 Overview In this course we will try to evelop a mathematical unerstaning for the

More information

Offloading Cellular Traffic through Opportunistic Communications: Analysis and Optimization

Offloading Cellular Traffic through Opportunistic Communications: Analysis and Optimization 1 Offloaing Cellular Traffic through Opportunistic Communications: Analysis an Optimization Vincenzo Sciancalepore, Domenico Giustiniano, Albert Banchs, Anreea Picu arxiv:1405.3548v1 [cs.ni] 14 May 24

More information

Robust PIM-SM Multicasting using Anycast RP in Wireless Ad Hoc Networks

Robust PIM-SM Multicasting using Anycast RP in Wireless Ad Hoc Networks Robust PIM-SM Multicasting using Anycast RP in Wireless A Hoc Networks Jaewon Kang, John Sucec, Vikram Kaul, Sunil Samtani an Mariusz A. Fecko Applie Research, Telcoria Technologies One Telcoria Drive,

More information

MORA: a Movement-Based Routing Algorithm for Vehicle Ad Hoc Networks

MORA: a Movement-Based Routing Algorithm for Vehicle Ad Hoc Networks : a Movement-Base Routing Algorithm for Vehicle A Hoc Networks Fabrizio Granelli, Senior Member, Giulia Boato, Member, an Dzmitry Kliazovich, Stuent Member Abstract Recent interest in car-to-car communications

More information

EDOVE: Energy and Depth Variance-Based Opportunistic Void Avoidance Scheme for Underwater Acoustic Sensor Networks

EDOVE: Energy and Depth Variance-Based Opportunistic Void Avoidance Scheme for Underwater Acoustic Sensor Networks sensors Article EDOVE: Energy an Depth Variance-Base Opportunistic Voi Avoiance Scheme for Unerwater Acoustic Sensor Networks Safar Hussain Bouk 1, *, Sye Hassan Ahme 2, Kyung-Joon Park 1 an Yongsoon Eun

More information

Questions? Post on piazza, or Radhika (radhika at eecs.berkeley) or Sameer (sa at berkeley)!

Questions? Post on piazza, or  Radhika (radhika at eecs.berkeley) or Sameer (sa at berkeley)! EE122 Fall 2013 HW3 Instructions Recor your answers in a file calle hw3.pf. Make sure to write your name an SID at the top of your assignment. For each problem, clearly inicate your final answer, bol an

More information

Generalized Edge Coloring for Channel Assignment in Wireless Networks

Generalized Edge Coloring for Channel Assignment in Wireless Networks TR-IIS-05-021 Generalize Ege Coloring for Channel Assignment in Wireless Networks Chun-Chen Hsu, Pangfeng Liu, Da-Wei Wang, Jan-Jan Wu December 2005 Technical Report No. TR-IIS-05-021 http://www.iis.sinica.eu.tw/lib/techreport/tr2005/tr05.html

More information

State Indexed Policy Search by Dynamic Programming. Abstract. 1. Introduction. 2. System parameterization. Charles DuHadway

State Indexed Policy Search by Dynamic Programming. Abstract. 1. Introduction. 2. System parameterization. Charles DuHadway State Inexe Policy Search by Dynamic Programming Charles DuHaway Yi Gu 5435537 503372 December 4, 2007 Abstract We consier the reinforcement learning problem of simultaneous trajectory-following an obstacle

More information

Learning convex bodies is hard

Learning convex bodies is hard Learning convex boies is har Navin Goyal Microsoft Research Inia navingo@microsoftcom Luis Raemacher Georgia Tech lraemac@ccgatecheu Abstract We show that learning a convex boy in R, given ranom samples

More information

AnyTraffic Labeled Routing

AnyTraffic Labeled Routing AnyTraffic Labele Routing Dimitri Papaimitriou 1, Pero Peroso 2, Davie Careglio 2 1 Alcatel-Lucent Bell, Antwerp, Belgium Email: imitri.papaimitriou@alcatel-lucent.com 2 Universitat Politècnica e Catalunya,

More information

Ad-Hoc Networks Beyond Unit Disk Graphs

Ad-Hoc Networks Beyond Unit Disk Graphs A-Hoc Networks Beyon Unit Disk Graphs Fabian Kuhn, Roger Wattenhofer, Aaron Zollinger Department of Computer Science ETH Zurich 8092 Zurich, Switzerlan {kuhn, wattenhofer, zollinger}@inf.ethz.ch ABSTRACT

More information

A Plane Tracker for AEC-automation Applications

A Plane Tracker for AEC-automation Applications A Plane Tracker for AEC-automation Applications Chen Feng *, an Vineet R. Kamat Department of Civil an Environmental Engineering, University of Michigan, Ann Arbor, USA * Corresponing author (cforrest@umich.eu)

More information

Skyline Community Search in Multi-valued Networks

Skyline Community Search in Multi-valued Networks Syline Community Search in Multi-value Networs Rong-Hua Li Beijing Institute of Technology Beijing, China lironghuascut@gmail.com Jeffrey Xu Yu Chinese University of Hong Kong Hong Kong, China yu@se.cuh.eu.h

More information

Non-homogeneous Generalization in Privacy Preserving Data Publishing

Non-homogeneous Generalization in Privacy Preserving Data Publishing Non-homogeneous Generalization in Privacy Preserving Data Publishing W. K. Wong, Nios Mamoulis an Davi W. Cheung Department of Computer Science, The University of Hong Kong Pofulam Roa, Hong Kong {wwong2,nios,cheung}@cs.hu.h

More information

Comparison of Methods for Increasing the Performance of a DUA Computation

Comparison of Methods for Increasing the Performance of a DUA Computation Comparison of Methos for Increasing the Performance of a DUA Computation Michael Behrisch, Daniel Krajzewicz, Peter Wagner an Yun-Pang Wang Institute of Transportation Systems, German Aerospace Center,

More information

Impact of FTP Application file size and TCP Variants on MANET Protocols Performance

Impact of FTP Application file size and TCP Variants on MANET Protocols Performance International Journal of Moern Communication Technologies & Research (IJMCTR) Impact of FTP Application file size an TCP Variants on MANET Protocols Performance Abelmuti Ahme Abbasher Ali, Dr.Amin Babkir

More information

Backpressure-based Packet-by-Packet Adaptive Routing in Communication Networks

Backpressure-based Packet-by-Packet Adaptive Routing in Communication Networks 1 Backpressure-base Packet-by-Packet Aaptive Routing in Communication Networks Eleftheria Athanasopoulou, Loc Bui, Tianxiong Ji, R. Srikant, an Alexaner Stolyar Abstract Backpressure-base aaptive routing

More information

Throughput Characterization of Node-based Scheduling in Multihop Wireless Networks: A Novel Application of the Gallai-Edmonds Structure Theorem

Throughput Characterization of Node-based Scheduling in Multihop Wireless Networks: A Novel Application of the Gallai-Edmonds Structure Theorem Throughput Characterization of Noe-base Scheuling in Multihop Wireless Networks: A Novel Application of the Gallai-Emons Structure Theorem Bo Ji an Yu Sang Dept. of Computer an Information Sciences Temple

More information

Multilevel Linear Dimensionality Reduction using Hypergraphs for Data Analysis

Multilevel Linear Dimensionality Reduction using Hypergraphs for Data Analysis Multilevel Linear Dimensionality Reuction using Hypergraphs for Data Analysis Haw-ren Fang Department of Computer Science an Engineering University of Minnesota; Minneapolis, MN 55455 hrfang@csumneu ABSTRACT

More information

1 Surprises in high dimensions

1 Surprises in high dimensions 1 Surprises in high imensions Our intuition about space is base on two an three imensions an can often be misleaing in high imensions. It is instructive to analyze the shape an properties of some basic

More information

A New Search Algorithm for Solving Symmetric Traveling Salesman Problem Based on Gravity

A New Search Algorithm for Solving Symmetric Traveling Salesman Problem Based on Gravity Worl Applie Sciences Journal 16 (10): 1387-1392, 2012 ISSN 1818-4952 IDOSI Publications, 2012 A New Search Algorithm for Solving Symmetric Traveling Salesman Problem Base on Gravity Aliasghar Rahmani Hosseinabai,

More information

Distributed Line Graphs: A Universal Technique for Designing DHTs Based on Arbitrary Regular Graphs

Distributed Line Graphs: A Universal Technique for Designing DHTs Based on Arbitrary Regular Graphs IEEE TRANSACTIONS ON KNOWLEDE AND DATA ENINEERIN, MANUSCRIPT ID Distribute Line raphs: A Universal Technique for Designing DHTs Base on Arbitrary Regular raphs Yiming Zhang an Ling Liu, Senior Member,

More information

Threshold Based Data Aggregation Algorithm To Detect Rainfall Induced Landslides

Threshold Based Data Aggregation Algorithm To Detect Rainfall Induced Landslides Threshol Base Data Aggregation Algorithm To Detect Rainfall Inuce Lanslies Maneesha V. Ramesh P. V. Ushakumari Department of Computer Science Department of Mathematics Amrita School of Engineering Amrita

More information

Characterizing Decoding Robustness under Parametric Channel Uncertainty

Characterizing Decoding Robustness under Parametric Channel Uncertainty Characterizing Decoing Robustness uner Parametric Channel Uncertainty Jay D. Wierer, Wahee U. Bajwa, Nigel Boston, an Robert D. Nowak Abstract This paper characterizes the robustness of ecoing uner parametric

More information

Transient analysis of wave propagation in 3D soil by using the scaled boundary finite element method

Transient analysis of wave propagation in 3D soil by using the scaled boundary finite element method Southern Cross University epublications@scu 23r Australasian Conference on the Mechanics of Structures an Materials 214 Transient analysis of wave propagation in 3D soil by using the scale bounary finite

More information

Loop Scheduling and Partitions for Hiding Memory Latencies

Loop Scheduling and Partitions for Hiding Memory Latencies Loop Scheuling an Partitions for Hiing Memory Latencies Fei Chen Ewin Hsing-Mean Sha Dept. of Computer Science an Engineering University of Notre Dame Notre Dame, IN 46556 Email: fchen,esha @cse.n.eu Tel:

More information

A Neural Network Model Based on Graph Matching and Annealing :Application to Hand-Written Digits Recognition

A Neural Network Model Based on Graph Matching and Annealing :Application to Hand-Written Digits Recognition ITERATIOAL JOURAL OF MATHEMATICS AD COMPUTERS I SIMULATIO A eural etwork Moel Base on Graph Matching an Annealing :Application to Han-Written Digits Recognition Kyunghee Lee Abstract We present a neural

More information

Coupon Recalculation for the GPS Authentication Scheme

Coupon Recalculation for the GPS Authentication Scheme Coupon Recalculation for the GPS Authentication Scheme Georg Hofferek an Johannes Wolkerstorfer Graz University of Technology, Institute for Applie Information Processing an Communications (IAIK), Inffelgasse

More information

Disjoint Multipath Routing in Dual Homing Networks using Colored Trees

Disjoint Multipath Routing in Dual Homing Networks using Colored Trees Disjoint Multipath Routing in Dual Homing Networks using Colore Trees Preetha Thulasiraman, Srinivasan Ramasubramanian, an Marwan Krunz Department of Electrical an Computer Engineering University of Arizona,

More information

Optimal Routing and Scheduling for Deterministic Delay Tolerant Networks

Optimal Routing and Scheduling for Deterministic Delay Tolerant Networks Optimal Routing an Scheuling for Deterministic Delay Tolerant Networks Davi Hay Dipartimento i Elettronica olitecnico i Torino, Italy Email: hay@tlc.polito.it aolo Giaccone Dipartimento i Elettronica olitecnico

More information

An Energy Efficient Routing for Wireless Sensor Networks: Hierarchical Approach

An Energy Efficient Routing for Wireless Sensor Networks: Hierarchical Approach An Energy Efficient Routing for Wireless Sensor Networks: Hierarchical Approach Nishi Sharma, Vanna Verma Abstract Wireless sensor networks (WSNs) is one of the emerging fiel of research in recent era

More information

Socially-optimal ISP-aware P2P Content Distribution via a Primal-Dual Approach

Socially-optimal ISP-aware P2P Content Distribution via a Primal-Dual Approach Socially-optimal ISP-aware P2P Content Distribution via a Primal-Dual Approach Jian Zhao, Chuan Wu The University of Hong Kong {jzhao,cwu}@cs.hku.hk Abstract Peer-to-peer (P2P) technology is popularly

More information

Inuence of Cross-Interferences on Blocked Loops: to know the precise gain brought by blocking. It is even dicult to determine for which problem

Inuence of Cross-Interferences on Blocked Loops: to know the precise gain brought by blocking. It is even dicult to determine for which problem Inuence of Cross-Interferences on Blocke Loops A Case Stuy with Matrix-Vector Multiply CHRISTINE FRICKER INRIA, France an OLIVIER TEMAM an WILLIAM JALBY University of Versailles, France State-of-the art

More information

Scalable Deterministic Scheduling for WDM Slot Switching Xhaul with Zero-Jitter

Scalable Deterministic Scheduling for WDM Slot Switching Xhaul with Zero-Jitter FDL sel. VOA SOA 100 Regular papers ONDM 2018 Scalable Deterministic Scheuling for WDM Slot Switching Xhaul with Zero-Jitter Bogan Uscumlic 1, Dominique Chiaroni 1, Brice Leclerc 1, Thierry Zami 2, Annie

More information

Improving Performance of Sparse Matrix-Vector Multiplication

Improving Performance of Sparse Matrix-Vector Multiplication Improving Performance of Sparse Matrix-Vector Multiplication Ali Pınar Michael T. Heath Department of Computer Science an Center of Simulation of Avance Rockets University of Illinois at Urbana-Champaign

More information

A shortest path algorithm in multimodal networks: a case study with time varying costs

A shortest path algorithm in multimodal networks: a case study with time varying costs A shortest path algorithm in multimoal networks: a case stuy with time varying costs Daniela Ambrosino*, Anna Sciomachen* * Department of Economics an Quantitative Methos (DIEM), University of Genoa Via

More information

Comparison of Wireless Network Simulators with Multihop Wireless Network Testbed in Corridor Environment

Comparison of Wireless Network Simulators with Multihop Wireless Network Testbed in Corridor Environment Comparison of Wireless Network Simulators with Multihop Wireless Network Testbe in Corrior Environment Rabiullah Khattak, Anna Chaltseva, Laurynas Riliskis, Ulf Boin, an Evgeny Osipov Department of Computer

More information

Politehnica University of Timisoara Mobile Computing, Sensors Network and Embedded Systems Laboratory. Testing Techniques

Politehnica University of Timisoara Mobile Computing, Sensors Network and Embedded Systems Laboratory. Testing Techniques Politehnica University of Timisoara Mobile Computing, Sensors Network an Embee Systems Laboratory ing Techniques What is testing? ing is the process of emonstrating that errors are not present. The purpose

More information

Online Appendix to: Generalizing Database Forensics

Online Appendix to: Generalizing Database Forensics Online Appenix to: Generalizing Database Forensics KYRIACOS E. PAVLOU an RICHARD T. SNODGRASS, University of Arizona This appenix presents a step-by-step iscussion of the forensic analysis protocol that

More information

Learning Subproblem Complexities in Distributed Branch and Bound

Learning Subproblem Complexities in Distributed Branch and Bound Learning Subproblem Complexities in Distribute Branch an Boun Lars Otten Department of Computer Science University of California, Irvine lotten@ics.uci.eu Rina Dechter Department of Computer Science University

More information

Blind Data Classification using Hyper-Dimensional Convex Polytopes

Blind Data Classification using Hyper-Dimensional Convex Polytopes Blin Data Classification using Hyper-Dimensional Convex Polytopes Brent T. McBrie an Gilbert L. Peterson Department of Electrical an Computer Engineering Air Force Institute of Technology 9 Hobson Way

More information

Additional Divide and Conquer Algorithms. Skipping from chapter 4: Quicksort Binary Search Binary Tree Traversal Matrix Multiplication

Additional Divide and Conquer Algorithms. Skipping from chapter 4: Quicksort Binary Search Binary Tree Traversal Matrix Multiplication Aitional Divie an Conquer Algorithms Skipping from chapter 4: Quicksort Binary Search Binary Tree Traversal Matrix Multiplication Divie an Conquer Closest Pair Let s revisit the closest pair problem. Last

More information

Learning Polynomial Functions. by Feature Construction

Learning Polynomial Functions. by Feature Construction I Proceeings of the Eighth International Workshop on Machine Learning Chicago, Illinois, June 27-29 1991 Learning Polynomial Functions by Feature Construction Richar S. Sutton GTE Laboratories Incorporate

More information

Backpressure-based Packet-by-Packet Adaptive Routing in Communication Networks

Backpressure-based Packet-by-Packet Adaptive Routing in Communication Networks 1 Backpressure-base Packet-by-Packet Aaptive Routing in Communication Networks Eleftheria Athanasopoulou, Loc Bui, Tianxiong Ji, R. Srikant, an Alexaner Stoylar arxiv:15.4984v1 [cs.ni] 27 May 21 Abstract

More information

Design of Policy-Aware Differentially Private Algorithms

Design of Policy-Aware Differentially Private Algorithms Design of Policy-Aware Differentially Private Algorithms Samuel Haney Due University Durham, NC, USA shaney@cs.ue.eu Ashwin Machanavajjhala Due University Durham, NC, USA ashwin@cs.ue.eu Bolin Ding Microsoft

More information

Estimating Velocity Fields on a Freeway from Low Resolution Video

Estimating Velocity Fields on a Freeway from Low Resolution Video Estimating Velocity Fiels on a Freeway from Low Resolution Vieo Young Cho Department of Statistics University of California, Berkeley Berkeley, CA 94720-3860 Email: young@stat.berkeley.eu John Rice Department

More information

Preamble. Singly linked lists. Collaboration policy and academic integrity. Getting help

Preamble. Singly linked lists. Collaboration policy and academic integrity. Getting help CS2110 Spring 2016 Assignment A. Linke Lists Due on the CMS by: See the CMS 1 Preamble Linke Lists This assignment begins our iscussions of structures. In this assignment, you will implement a structure

More information

Control of Scalable Wet SMA Actuator Arrays

Control of Scalable Wet SMA Actuator Arrays Proceeings of the 2005 IEEE International Conference on Robotics an Automation Barcelona, Spain, April 2005 Control of Scalable Wet SMA Actuator Arrays eslie Flemming orth Dakota State University Mechanical

More information

An Investigation in the Use of Vehicle Reidentification for Deriving Travel Time and Travel Time Distributions

An Investigation in the Use of Vehicle Reidentification for Deriving Travel Time and Travel Time Distributions An Investigation in the Use of Vehicle Reientification for Deriving Travel Time an Travel Time Distributions Carlos Sun Department of Civil an Environmental Engineering, University of Missouri-Columbia,

More information

Non-Uniform Sensor Deployment in Mobile Wireless Sensor Networks

Non-Uniform Sensor Deployment in Mobile Wireless Sensor Networks 01 01 01 01 01 00 01 01 Non-Uniform Sensor Deployment in Mobile Wireless Sensor Networks Mihaela Carei, Yinying Yang, an Jie Wu Department of Computer Science an Engineering Floria Atlantic University

More information

Study of Network Optimization Method Based on ACL

Study of Network Optimization Method Based on ACL Available online at www.scienceirect.com Proceia Engineering 5 (20) 3959 3963 Avance in Control Engineering an Information Science Stuy of Network Optimization Metho Base on ACL Liu Zhian * Department

More information

Using Vector and Raster-Based Techniques in Categorical Map Generalization

Using Vector and Raster-Based Techniques in Categorical Map Generalization Thir ICA Workshop on Progress in Automate Map Generalization, Ottawa, 12-14 August 1999 1 Using Vector an Raster-Base Techniques in Categorical Map Generalization Beat Peter an Robert Weibel Department

More information

Coupon Recalculation for the GPS Authentication Scheme

Coupon Recalculation for the GPS Authentication Scheme Coupon Recalculation for the GPS Authentication Scheme Georg Hofferek an Johannes Wolkerstorfer Graz University of Technology, Institute for Applie Information Processing an Communications (IAIK), Inffelgasse

More information

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 31, NO. 4, APRIL

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 31, NO. 4, APRIL IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 1, NO. 4, APRIL 01 74 Towar Efficient Distribute Algorithms for In-Network Binary Operator Tree Placement in Wireless Sensor Networks Zongqing Lu,

More information

Module13:Interference-I Lecture 13: Interference-I

Module13:Interference-I Lecture 13: Interference-I Moule3:Interference-I Lecture 3: Interference-I Consier a situation where we superpose two waves. Naively, we woul expect the intensity (energy ensity or flux) of the resultant to be the sum of the iniviual

More information

New Version of Davies-Bouldin Index for Clustering Validation Based on Cylindrical Distance

New Version of Davies-Bouldin Index for Clustering Validation Based on Cylindrical Distance New Version of Davies-Boulin Inex for lustering Valiation Base on ylinrical Distance Juan arlos Roas Thomas Faculta e Informática Universia omplutense e Mari Mari, España correoroas@gmail.com Abstract

More information

Uninformed search methods

Uninformed search methods CS 1571 Introuction to AI Lecture 4 Uninforme search methos Milos Hauskrecht milos@cs.pitt.eu 539 Sennott Square Announcements Homework assignment 1 is out Due on Thursay, September 11, 014 before the

More information

Data Mining: Clustering

Data Mining: Clustering Bi-Clustering COMP 790-90 Seminar Spring 011 Data Mining: Clustering k t 1 K-means clustering minimizes Where ist ( x, c i t i c t ) ist ( x m j 1 ( x ij i, c c t ) tj ) Clustering by Pattern Similarity

More information

An Algorithm for Building an Enterprise Network Topology Using Widespread Data Sources

An Algorithm for Building an Enterprise Network Topology Using Widespread Data Sources An Algorithm for Builing an Enterprise Network Topology Using Wiesprea Data Sources Anton Anreev, Iurii Bogoiavlenskii Petrozavosk State University Petrozavosk, Russia {anreev, ybgv}@cs.petrsu.ru Abstract

More information

Multi-camera tracking algorithm study based on information fusion

Multi-camera tracking algorithm study based on information fusion International Conference on Avance Electronic Science an Technolog (AEST 016) Multi-camera tracking algorithm stu base on information fusion a Guoqiang Wang, Shangfu Li an Xue Wen School of Electronic

More information

An Infrastructureless End-to-End High Performance Mobility Protocol

An Infrastructureless End-to-End High Performance Mobility Protocol Proceeings of the 5 IEEE International Conference on Electro Information Technology, 5, Lincoln Nebraska, May 5 An Infrastructureless En-to-En High Performance Mobility Protocol Saneep Davu, Rai Y. Zaghal

More information

Implementation and Evaluation of NAS Parallel CG Benchmark on GPU Cluster with Proprietary Interconnect TCA

Implementation and Evaluation of NAS Parallel CG Benchmark on GPU Cluster with Proprietary Interconnect TCA Implementation an Evaluation of AS Parallel CG Benchmark on GPU Cluster with Proprietary Interconnect TCA Kazuya Matsumoto 1, orihisa Fujita 2, Toshihiro Hanawa 3, an Taisuke Boku 1,2 1 Center for Computational

More information

SURVIVABLE IP OVER WDM: GUARANTEEEING MINIMUM NETWORK BANDWIDTH

SURVIVABLE IP OVER WDM: GUARANTEEEING MINIMUM NETWORK BANDWIDTH SURVIVABLE IP OVER WDM: GUARANTEEEING MINIMUM NETWORK BANDWIDTH Galen H Sasaki Dept Elec Engg, U Hawaii 2540 Dole Street Honolul HI 96822 USA Ching-Fong Su Fuitsu Laboratories of America 595 Lawrence Expressway

More information

Data Mining: Concepts and Techniques. Chapter 7. Cluster Analysis. Examples of Clustering Applications. What is Cluster Analysis?

Data Mining: Concepts and Techniques. Chapter 7. Cluster Analysis. Examples of Clustering Applications. What is Cluster Analysis? Data Mining: Concepts an Techniques Chapter Jiawei Han Department of Computer Science University of Illinois at Urbana-Champaign www.cs.uiuc.eu/~hanj Jiawei Han an Micheline Kamber, All rights reserve

More information

Recitation Caches and Blocking. 4 March 2019

Recitation Caches and Blocking. 4 March 2019 15-213 Recitation Caches an Blocking 4 March 2019 Agena Reminers Revisiting Cache Lab Caching Review Blocking to reuce cache misses Cache alignment Reminers Due Dates Cache Lab (Thursay 3/7) Miterm Exam

More information

CS269I: Incentives in Computer Science Lecture #8: Incentives in BGP Routing

CS269I: Incentives in Computer Science Lecture #8: Incentives in BGP Routing CS269I: Incentives in Computer Science Lecture #8: Incentives in BGP Routing Tim Roughgaren October 19, 2016 1 Routing in the Internet Last lecture we talke about elay-base (or selfish ) routing, which

More information

Adjacency Matrix Based Full-Text Indexing Models

Adjacency Matrix Based Full-Text Indexing Models 1000-9825/2002/13(10)1933-10 2002 Journal of Software Vol.13, No.10 Ajacency Matrix Base Full-Text Inexing Moels ZHOU Shui-geng 1, HU Yun-fa 2, GUAN Ji-hong 3 1 (Department of Computer Science an Engineering,

More information

Adaptive Load Balancing based on IP Fast Reroute to Avoid Congestion Hot-spots

Adaptive Load Balancing based on IP Fast Reroute to Avoid Congestion Hot-spots Aaptive Loa Balancing base on IP Fast Reroute to Avoi Congestion Hot-spots Masaki Hara an Takuya Yoshihiro Faculty of Systems Engineering, Wakayama University 930 Sakaeani, Wakayama, 640-8510, Japan Email:

More information

CS350 - Exam 4 (100 Points)

CS350 - Exam 4 (100 Points) Fall 0 Name CS0 - Exam (00 Points).(0 points) Re-Black Trees For the Re-Black tree below, inicate where insert() woul initially occur before rebalancing/recoloring. Sketch the tree at ALL intermeiate steps

More information

Random Clustering for Multiple Sampling Units to Speed Up Run-time Sample Generation

Random Clustering for Multiple Sampling Units to Speed Up Run-time Sample Generation DEIM Forum 2018 I4-4 Abstract Ranom Clustering for Multiple Sampling Units to Spee Up Run-time Sample Generation uzuru OKAJIMA an Koichi MARUAMA NEC Solution Innovators, Lt. 1-18-7 Shinkiba, Koto-ku, Tokyo,

More information

Research Article REALFLOW: Reliable Real-Time Flooding-Based Routing Protocol for Industrial Wireless Sensor Networks

Research Article REALFLOW: Reliable Real-Time Flooding-Based Routing Protocol for Industrial Wireless Sensor Networks Hinawi Publishing Corporation International Journal of Distribute Sensor Networks Volume 2014, Article ID 936379, 17 pages http://x.oi.org/10.1155/2014/936379 Research Article REALFLOW: Reliable Real-Time

More information

Using Ray Tracing for Site-Specific Indoor Radio Signal Strength Analysis 1

Using Ray Tracing for Site-Specific Indoor Radio Signal Strength Analysis 1 Using Ray Tracing for Site-Specific Inoor Raio Signal Strength Analysis 1 Michael Ni, Stephen Mann, an Jay Black Computer Science Department, University of Waterloo, Waterloo, Ontario, NL G1, Canaa Abstract

More information

Performance evaluation of the Zipper duplex method

Performance evaluation of the Zipper duplex method Performance evaluation of the Zipper lex metho Frank Sjöberg, Mikael Isaksson, Petra Deutgen, Rickar Nilsson, Per Öling, an Per Ola Börjesson Luleå University of Technology, Division of Signal Processing,

More information

Feature Extraction and Rule Classification Algorithm of Digital Mammography based on Rough Set Theory

Feature Extraction and Rule Classification Algorithm of Digital Mammography based on Rough Set Theory Feature Extraction an Rule Classification Algorithm of Digital Mammography base on Rough Set Theory Aboul Ella Hassanien Jafar M. H. Ali. Kuwait University, Faculty of Aministrative Science, Quantitative

More information

All-to-all Broadcast for Vehicular Networks Based on Coded Slotted ALOHA

All-to-all Broadcast for Vehicular Networks Based on Coded Slotted ALOHA Preprint, August 5, 2018. 1 All-to-all Broacast for Vehicular Networks Base on Coe Slotte ALOHA Mikhail Ivanov, Frerik Brännström, Alexanre Graell i Amat, an Petar Popovski Department of Signals an Systems,

More information

Solution Representation for Job Shop Scheduling Problems in Ant Colony Optimisation

Solution Representation for Job Shop Scheduling Problems in Ant Colony Optimisation Solution Representation for Job Shop Scheuling Problems in Ant Colony Optimisation James Montgomery, Carole Faya 2, an Sana Petrovic 2 Faculty of Information & Communication Technologies, Swinburne University

More information

CS 106 Winter 2016 Craig S. Kaplan. Module 01 Processing Recap. Topics

CS 106 Winter 2016 Craig S. Kaplan. Module 01 Processing Recap. Topics CS 106 Winter 2016 Craig S. Kaplan Moule 01 Processing Recap Topics The basic parts of speech in a Processing program Scope Review of syntax for classes an objects Reaings Your CS 105 notes Learning Processing,

More information

Yet Another Parallel Hypothesis Search for Inverse Entailment Hiroyuki Nishiyama and Hayato Ohwada Faculty of Sci. and Tech. Tokyo University of Scien

Yet Another Parallel Hypothesis Search for Inverse Entailment Hiroyuki Nishiyama and Hayato Ohwada Faculty of Sci. and Tech. Tokyo University of Scien Yet Another Parallel Hypothesis Search for Inverse Entailment Hiroyuki Nishiyama an Hayato Ohwaa Faculty of Sci. an Tech. Tokyo University of Science, 2641 Yamazaki, Noa-shi, CHIBA, 278-8510, Japan hiroyuki@rs.noa.tus.ac.jp,

More information

MODULE VII. Emerging Technologies

MODULE VII. Emerging Technologies MODULE VII Emerging Technologies Computer Networks an Internets -- Moule 7 1 Spring, 2014 Copyright 2014. All rights reserve. Topics Software Define Networking The Internet Of Things Other trens in networking

More information

Figure 1: Schematic of an SEM [source: ]

Figure 1: Schematic of an SEM [source:   ] EECI Course: -9 May 1 by R. Sanfelice Hybri Control Systems Eelco van Horssen E.P.v.Horssen@tue.nl Project: Scanning Electron Microscopy Introuction In Scanning Electron Microscopy (SEM) a (bunle) beam

More information

Supporting Fully Adaptive Routing in InfiniBand Networks

Supporting Fully Adaptive Routing in InfiniBand Networks XIV JORNADAS DE PARALELISMO - LEGANES, SEPTIEMBRE 200 1 Supporting Fully Aaptive Routing in InfiniBan Networks J.C. Martínez, J. Flich, A. Robles, P. López an J. Duato Resumen InfiniBan is a new stanar

More information

Secure Network Coding for Distributed Secret Sharing with Low Communication Cost

Secure Network Coding for Distributed Secret Sharing with Low Communication Cost Secure Network Coing for Distribute Secret Sharing with Low Communication Cost Nihar B. Shah, K. V. Rashmi an Kannan Ramchanran, Fellow, IEEE Abstract Shamir s (n,k) threshol secret sharing is an important

More information