Cooperative Data Dissemination to Mission Sites
|
|
- Nelson Clarke
- 5 years ago
- Views:
Transcription
1 Cooperative Data Dissemination to Mission Sites Fangfei Chen a, Matthew P. Johnson b, Amotz Bar-Noy b and Thomas F. La Porta a a Department of Computer Science and Engineering, the Pennsylvania State University b Department of Computer Science, the City University of New York Graduate Center ABSTRACT Timely dissemination of information to mobile users is vital in many applications. In a critical situation, no network infrastructure may be available for use in dissemination, over and above the on-board storage capability of the mobile users themselves. We consider the following specialized content distribution application: a group of users equipped with wireless devices build an ad hoc network in order cooperatively to retrieve information from certain regions (the mission sites). Each user requires access to some set of information items originating from sources lying within a region. Each user desires low-latency access to its desired data items, upon request (i.e., when pulled). In order to minimize average response time, we allow users to pull data either directly from sources or, when possible, from other nearby users who have already pulled, and continue to carry, the desired data items. That is, we allow for data to be pushed to one user and then pulled by one or more additional users. The total latency experienced by a user vis-vis a certain data item is then in general a combination of the push delay and the pull delay. We assume each delay time is a function of the hop distance between the pair of points in question. Our goal in this paper is to assign data to mobile users, in order to minimize the total cost and the average latency experienced by all the users. In a static setting, we solve this problem in two different schemes, one of which is easy to solve but wasteful, one of which relates to NP-hard problems but is less so. Then in a dynamic setting, we adapt the algorithm for the static setting and develop a new algorithm with respect to users gradual arrival. In the end we show a trade-off can be made between minimizing the cost and latency. Introduction Timely dissemination of information to mobile users is vital in many applications. In a critical situation, no network infrastructure may be available for use in dissemination, over and above the on-board storage capability of the mobile users themselves. We consider the following specialized content distribution application: a group of users equipped with wireless devices build an ad hoc network in order cooperatively to retrieve information from certain regions (the mission sites). Each user requires access to some set of information items originating from sources lying throughout a region. Each user desires low-latency access to its desired data items, upon request (i.e., when pulled). In order to minimize average response time, we allow users to pull data either directly from sources or, when possible, from other nearby users who have already pulled, and continue to carry, the desired data items. That is, we allow for data to be pushed to one user and then pulled by one or more additional users. In previous work, for a similar scenario, data is pushed to some storage nodes at the mission site where it may be retrieved locally when users arrive. The goal is to choose a storage node assignment minimizing the total latency-based cost. Since the number and storage capacity of storage nodes are both limited, the problem is proved to be NP-hard even to approximate. In this work, we assume no network infrastructure is available for assistance, that is, the group of users need to build an ad hoc network and cooperatively retrieve information by themselves. We start with a static problem setting in which all users already arrive when the dissemination starts. Users not within the communication range of each other can communicate by multi-hop routing protocols. Assuming the network is connected, data can be delivered to any of the users. Our goal in this setting is to minimize the total number of hops it takes to disseminate every data to the requesting users. We assume each user has enough space to contribute in the dissemination, which makes a simple sharing scheme, One-to-Many, solvable in polynomial time. However, when we try to further reduce the cost through a more complicated sharing scheme, computing the costs involves solving NP-hard problems. We provide a heuristic approximation algorithm for this setting and compare its performance, through simulation, with the One-to-Many Scheme s.
2 Then we introduce the dynamic setting, that is, users gradually arrive after the dissemination begins. Besides the number of hops, we take user-experienced latency into consideration. That is, the time between when users arrive and when they receive the first piece of their data. Since we have no dedicated storage nodes, data can be retrieved only from users that have already arrived. Another complication is that the network may not be fully connected until a certain point. Therefore, always trying to minimize the number of hops as in the static setting may increase the average latency. We adapt our algorithm for this dynamic setting as a base-level comparison and propose a new greedy algorithm. Finally, we demonstrate a trade-off between minimizing the number of hops and the average latency. The rest of this paper is organized as follows. Section presents related work. Section formulates the problem in static settings and propose two schemes for solving the problem. Section shows how we generate simulations and evaluate the algorithms. Section 5 discusses challenges encountered when solving the problem in the dynamic setting and proposes a new algorithm. Section 6 concludes the paper. Related Work Our efforts here relate to a body of work on Peer-to-Peer Content Distribution (PP CDN). Such a system, according to the survey of Androutsellis-Theotokis and Spinellis, creates a distributed storage medium that allows for the publishing, searching, and retrieval of files by members of its network. Unlike a traditional CDN, a PP Content Distribution Network requires very few or even no network infrastructure and is therefore scalable and easy to deploy. Our model differs from PP CDNs, however, in that the information needs for each user are predefined. PP CDNs are designed to provide users access to all available files, in which case files are stored and indexed that may not be requested over long periods of time. We are motivated by critical situations in which all available bandwidth must be taken advantage of in order to serve as many user demands as possible. Therefore we store and deliver only data for which there is need. Another more technical difference is that each data item in our problem is treated as an unsplittable whole, different copies of which appear for different users, while in PP CDNs data is normally broken into parts and stored in a distributed way. Another important aspect of our work is that each data item can be requested by multiple users, which relates the problem to the study of multicast protocols. The benefit of multicast over using traditional multiple unicast links is that the source only needs to do a single transmission of the data, which will then be propagated by the network. Analogously, in our work sources send data items to one user, who then makes the item available to other users who in turn may share it further. One specific scenario we study is one in which users gradually arrive; one of our considerations in evaluating solutions is user-experienced latency. This paper is a direct extension of our previous work, in which data is pushed to storage nodes at the mission site where it may be retrieved locally when users arrive. The storage space on each nearby node is limited, yielding a variant of the Multiple Knapsack Problem in which our goal is to choose a storage node assignment minimizing the total latency-based cost. We used a simple One-to-Many Scheme, under which all users need to retrieve data from the storage nodes. In this new work, we attempt to improve the one-to-many dissemination structure by introducing a spanning tree structure or even a dynamically growing structure. Each data item, newly delivered to its first user, presents an instance of the Steiner Tree problem, which is a classical NP-hard problem 5 originally posed by Gauss. That problem is known to admit a.55-approximation, 6 but a simple -approximation algorithm 7 based on spanning trees is well known, which we adapt here as a heuristic algorithm. In the previous work we used a static network topology, assuming storages nodes were deployed ahead of time, with fixed locations, so users experienced latency could vary depending on whether the data was ready at the storage node at the time of users arrival. In the current paper, however, the latency varies mainly because of changes to network topology over time. Evaluating different strategies to deal with this latency, we show that a trade-off can be achieved between latency and cost. Static Settings We start with problem settings in which all mobile users have already arrived at the mission site when the dissemination begins. In this section, we formally define the problem and discuss two solution schemes, which we evaluate through simulation later in the paper.
3 . Problem definition Let D = {d i : i =... m} be a set of data items and U = {u j : j =... n} be a set of users. Each data item d i has a size s i and is co-located with one of the users (we need at least one user who can reach the data item in only one hop). Each user must retrieve some subset D j D of the data items. Unlike in the previous work, we assume each user has enough storage space for both storing its own items and passing on others items. Each potential placement of d i in u j (u j does not necessarily request d j ) is associated with some cost c ij. The objective is to minimize the cost of dissemination of all data items. The problem is formally defined as: min i,j c ij x ij () s.t. j x ij = i x ij {0, } i, j Hop count is a natural metric for cost of delivering data in wireless networks. In previous work, we used distance as a proxy of hop count. In this work, we more accurately use the exact number of hops based on the shortest path routing. It is crucial to note that different dissemination schemes will result in different cost values c ij, and indeed computing the true values of these costs may not be easy.. One-to-Many Scheme A naive approach to solve this new problem is to directly apply the previous algorithm (for the problem setting in which items are assigned to one storage node and then pulled by users interested). We call it One-to- Many scheme (see Figure (a)). In this scheme, one user stores the data item and then other users who request the data item retrieve it only from this first user. In this way, the user simulates the functionality of a storage node in the previous work. The cost c ij of placing data item j in user i is based on the total cost for all other users desiring item j to access it direct from user i. Let U i denote the subset of users requesting data item i, (a) One-to-Many Scheme. (b) Spanning Tree Scheme. Figure : Two Schemes. and let h xy be the minimum number of hops from u x to u y. Then the cost to store data i on user j is: c ij = s i (h ij + ) () k:u k U i h jk Furthermore, we may consider the user who has direct access to d i as another user requesting d i, then we may remove the push part from the calculation: c ij = s i h jk () k:u k U i
4 h xy can be obtained by Dijkstra s algorithm in polynomial time. Since there is no capacity constraint on the users, the optimal placement decisions for the data items can all be made independently, by enumeration in polynomial time. Given the specified cost values c ij, the One-to-Many Scheme (see Algorithm ), chooses the best placement for each item. Algorithm One-to-Many Scheme : for i = to m do : for j = to n do : c ij s i k:u k U i h jk : end for 5: j argmin c ij j 6: assign d i to u j 7: end for. Spanning Tree Scheme Although the One-to-Many Scheme is very simple, it likely introduces a lot of redundant traffic between the user who stores the data and other users. A more sophisticated approach is to use an algorithm that takes advantage of all the freedom this new problem setting affords (i.e., allowing each user to pass the item to the next). Since data only needs to reach each recipient once, what is desired for each data item is a tree structure rooted at the first recipient and including all the other recipients as nodes (see Figure (b)). The nodes of a given data item s tree need not be limited to users desiring that item, however, since we allow users to carry items that they do not themselves need, in order to provide aid users. The tree for each data item therefore is required to contain nodes for all that item s users, the node who has direct access to it and it may optionally include nodes for other users as well. The cost of a tree is the sum of its edge-weights. Finding the lowest-cost such tree is an instance of the Steiner tree 5 problem: given a graph G = (V, E) and a subset S V, find a minimum-cost connected subgraph of G that spans all vertices of S. Although the Steiner Tree problem is NP-hard in general, it admits a.55-approximation algorithm. 6 There is also a well known factor -approximation algorithm (discovered by many authors; see Vazirani 7 ) that begins by computing a minimum spanning tree (MST) on the graph induced by the required nodes S. We opt to use a variation on the MST-based algorithm which, while sacrificing the approximation guarantee, has practical advantages in efficiency. We compute a spanning tree out of G for each user j and construct it as being rooted at j. For each item i that we consider placing in user j, we prune any redundant leaves from the tree rooted at j, i.e., users who do not desire item j and are not needed to forward it to others. User j can then forward any data items it possesses to users desiring them, along the paths of this tree. We use Algorithm to calculate Algorithm Tree Scheme calculation of c ij values : T = T (V, E) is the MST rooted at u j : V V, E E : sort vertices V in level order in T : for k = n down to do 5: if v k is a leaf node and u k / U i then 6: remove v k from V 7: remove all edges of v k from E 8: end if 9: c ij s i E 0: end for the actual number of hops implied by the pruned spanning tree. Given the computed c ij values, we again use Algorithm (except with the new computation of c ij replacing line ) to choose the best assignment for each item. As assignment is now understood to mean that item i will be disseminated from the pruned spanning tree rooted at user j.
5 Evaluation In this section, we evaluate the two algorithms under different circumstances. First we present the evaluation settings and explain how we generate random problem instances. Then we explain the results and compare the algorithms.. Problem instance generation We use undirected graphs to model the network of communication between users. For each set of parameters we test, we first need to generate a random graph representing the problem instance. We require that the graph is connected so that there is a route from any user to any other user in the network. For this, we use a modified version of the Barabasi and Albert scale-free network model 8 (see Algorithm ). In the original paper, 8 the Algorithm Random Connected Graph Generation : G = {V, E} = : for j = to n do : V V v j, nedges = 0 : for i = to j do 5: r random() 6: prob C deg(v i )/ E 7: if r prob then 8: E E {(v i, v j )} 9: nedges nedges + 0: end if : end for : if nedges = 0 then : v j has the largest deg(v j ) : E E {(v i, v j )} 5: end if 6: end for probability p of creating an edge between an existing vertex v and the newly added vertex is: p = deg(v)/ E Here deg(v) indicates the degree of vertex v, i.e., the number of users v can directly communicate with. We multiply this probability by a factor of C (the vertex degree parameter), in order to adjust the number of edges in the graph. Each time we add a new vertex, we check the number of edges we add connecting to this vertex. If no edge is added, we simply connect the new vertex to the vertex with the largest degree so that the graph remains connected after each step. The sizes of data items are selected uniformly at random from [, maxdatasize]. Assuming at least one user can read each data item directly, each data item is randomly assigned to one of the users as sources. Each data item is requested by a given user with probability reqp rob. By default, we set the number of data items to 50, the number of users to 50, reqp rob = 0., C = and maxdatasize = 0. To test the performance of the algorithms, we count the total number of hops in four separate series of simulations, each varying one of the following parameters: the number of the data items, the number of users, reqp rob and C.. Results The Spanning Tree scheme constantly beats the One-to-Many Scheme in these four tests (see Figure ), although their divergence is not unbounded. When we expand the problem instance size, by increasing either the number of data items or the number of users, we consistently find One-to-Many Scheme s hop count to be within a constant factor of Spanning Tree s (see Figures (a) and (b)).
6 8 x 0 7 One to Many Scheme Spanning Tree Scheme x 0 One to Many Scheme Spanning Tree Scheme 6 0 Number of hops 5 Number of hops Number of data items (a) The number of data items Number of users (b) The number of users. 8 x 0 7 One to Many Scheme Spanning Tree Scheme.5 x 05. One to Many Scheme Spanning Tree Scheme 6.5 Number of hops 5 Number of hops User request rate (c) The user request rate Vertex degree parameter (d) The vertex degree parameter. Figure : Evaluation results, varying the specified parameter. In the third test, we fix the numbers of data items and users while varying the user request probability reqp rob. As we increase reqp rob, each data item is requested by more and more users. In Figure (c), we see that the number of hops by the One-to-Many Scheme increases linearly, because each additional request will result in a new path from the root that stores the data. The hop count of the Spanning Tree Scheme tends to grow sublinearly, because we typically only need to extend the existing tree structure modestly for each additional request. In the fourth test, we increase the probability with which a new vertex connects to existing nodes during the random graph generation. As the probability increases average degree, so the depth of the graph decreases. Therefore, we expect the number of hops to decrease for both the two schemes, which is indeed what we observe (see Figure (d)). 5 Dynamic Settings In this section, we start to look at the case in which users arrive gradually after the dissemination begins. First, we adapt our algorithms for the static case to this new setting. Then we propose a greedy algorithm which takes into consideration the user-experienced latency. We again run simulations to compare these algorithms. Since users arrive at different moments in this dynamic setting, they may experience highly variable retrieval latency even for the same data item. Because the network connectivity changes over time, some users may even be unreachable until a certain point. As a result, the time users spend waiting for connecting to the network may be much longer than the actual transmission latency. We define user-experienced latency to be the time between a user s arrival and the moment when it receives the first part of the data item.
7 5. Spanning Tree The Spanning Tree Scheme focuses on minimizing the total number of hops; it takes no consideration of the users arrival time. Moreover, it assumes the current presence of all nodes. Nonetheless, we can apply it here by ) assuming a priori knowledge of nodes that will eventually arrive, and ) disseminating via the spanning tree and caching data as needed until each next-hop node arrives. Once the nodes arrive, we can continue with the dissemination immediately. When a user arrives, however, he may have to wait until all his ancestor nodes in the spanning tree have also arrived. 5. Greedy A trade-off can be made if we allow the dissemination not to follow the spanning tree. Then, as soon as a user becomes reachable from other users, we can deliver all its requested data items following the current shortest paths, which may be different from the shortest paths in the full graph. We call this the Greedy algorithm (see Algorithm ). Algorithm Greedy algorithm : for each new user arrival do : update the connectivity graph G : for each user j that becomes newly reachable do : for each data item i that user j requests do 5: find the nearest user k that has data i in G 6: deliver data i from user k to user j 7: end for 8: end for 9: end for 5. Evaluation We compare the two algorithms in terms of the total number of hops and the sum of users experienced latency. For simplicity, we assume that one unit size data needs one time slot to do one hop transmission. We conduct simulations similar to those in Section, except that we record the user-experienced latency as well. 8 x 05 7 Greedy Spanning Tree.5 x 05 Greedy Spanning Tree Total experienced delay 6 5 Total number of hops Number of Data Items (a) Total experienced delay Number of Data Items (b) Total number of hops. Figure : Trials varying the number of data items. In the first set of experiments, we vary the number of data items. In Figure (a), we can see that the Greedy algorithm always results in lower latency than the Spanning Tree algorithm. In Figure (b), we see that the spanning tree always beats the Greedy algorithm in terms of number of hops. This is what we expected based on the focuses of the two algorithms. In the second set of experiments, we increase the vertex degree factor. For the Greedy algorithm (see Figures (a) and (b)), we see both delay and number of hops decreasing quickly, because increasing the degree of
8 .5 x 05 Greedy Spanning Tree.7 x 05.6 Greedy Spanning Tree Total experienced delay.5.5 Total number of hops Vertex degree factor (a) Total experienced delay. Figure : Trials varying the vertex degree factor Vertex degree factor (b) Total number of hops. each node will both help the graph become connected sooner tend to decrease path lengths. The delay for the Spanning Tree algorithm, however, hardly decreases. Although nodes become connected to the graph sooner, since we force them to follow the spanning tree, the experienced latency is mostly determined by the parent node who arrives the last. The number of hops decreases as also happened in the former experiments. 6 Conclusion In this paper, we have studied a scenario in which users build an ad-hoc network to cooperatively disseminate data without relying on any network infrastructure. We start with a scenario in which all users arrive before the dissemination begins. Then we adopt our prior algorithms and propose a new algorithm for the case in which users arrive gradually over time. We evaluate these algorithms under different circumstances and show that a trade-off can be made between the total number of hops and the user-experienced latencies. ACKNOWLEDGMENTS Research was sponsored by US Army Research laboratory and the UK Ministry of Defence and was accomplished under Agreement Number W9NF The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the US Army Research Laboratory, the U.S. Government, the UK Ministry of Defense, or the UK Government. The US and UK Governments are authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation hereon. REFERENCES. F. Chen, M. P. Johnson, A. Bar-Noy, I. Fermin, and T. La Porta, Proactive data dissemination to mission sites, 6th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks, June S. Androutsellis-Theotokis and D. Spinellis, A survey of peer-to-peer content distribution technologies, ACM Computing Surveys (CSUR) 6(), p. 7, 00.. S. Deering and D. Cheriton, Multicast routing in datagram internetworks and extended LANs, ACM Transactions on Computer Systems (TOCS) 8(), pp. 85 0, H. Kellerer, U. Pferschy, and D. Pisinger, Knapsack Problems, Springer, M. Garey and D. Johnson, Computers and Intractability: A Guide to the Theory of NP-Colmpleteness, Freeman, G. Robins and A. Zelikovsky, Tighter bounds for graph steiner tree approximation, SIAM J. Discrete Math. 9(), pp., V. V. Vazirani, Approximation Algorithms, Springer, A. Barabasi and R. Albert, Emergence of scaling in random networks, Science 86(59), p. 509, 999.
Utility-Based Joint Sensor Selection and Congestion Control for Task-Oriented WSNs
Utility-Based Joint Sensor Selection and Congestion Control for Task-Oriented WSNs Hosam Rowaihy, Matthew P. Johnson, Sharanya Eswaran Diego Pizzocaro, Amotz Bar-Noy, Thomas La Porta, Archan Misra and
More informationApproximating Node-Weighted Multicast Trees in Wireless Ad-Hoc Networks
Approximating Node-Weighted Multicast Trees in Wireless Ad-Hoc Networks Thomas Erlebach Department of Computer Science University of Leicester, UK te17@mcs.le.ac.uk Ambreen Shahnaz Department of Computer
More informationApplication Layer Multicast Algorithm
Application Layer Multicast Algorithm Sergio Machado Universitat Politècnica de Catalunya Castelldefels Javier Ozón Universitat Politècnica de Catalunya Castelldefels Abstract This paper presents a multicast
More informationCONSTRUCTION AND EVALUATION OF MESHES BASED ON SHORTEST PATH TREE VS. STEINER TREE FOR MULTICAST ROUTING IN MOBILE AD HOC NETWORKS
CONSTRUCTION AND EVALUATION OF MESHES BASED ON SHORTEST PATH TREE VS. STEINER TREE FOR MULTICAST ROUTING IN MOBILE AD HOC NETWORKS 1 JAMES SIMS, 2 NATARAJAN MEGHANATHAN 1 Undergrad Student, Department
More information6% Writes. 4% Writes. 2% Writes. D e g r e e o f R e p l i c a t i o n
Power-Aware Replication of Data Structures in Distributed Embedded Real-Time Systems? Osman S. Unsal, Israel Koren, C. Mani Krishna Department of Electrical and Computer Engineering University of Massachusetts,
More informationON WEIGHTED RECTANGLE PACKING WITH LARGE RESOURCES*
ON WEIGHTED RECTANGLE PACKING WITH LARGE RESOURCES* Aleksei V. Fishkin, 1 Olga Gerber, 1 Klaus Jansen 1 1 University of Kiel Olshausenstr. 40, 24118 Kiel, Germany {avf,oge,kj}@informatik.uni-kiel.de Abstract
More informationCommunication Networks I December 4, 2001 Agenda Graph theory notation Trees Shortest path algorithms Distributed, asynchronous algorithms Page 1
Communication Networks I December, Agenda Graph theory notation Trees Shortest path algorithms Distributed, asynchronous algorithms Page Communication Networks I December, Notation G = (V,E) denotes a
More information3 No-Wait Job Shops with Variable Processing Times
3 No-Wait Job Shops with Variable Processing Times In this chapter we assume that, on top of the classical no-wait job shop setting, we are given a set of processing times for each operation. We may select
More informationDistributed minimum spanning tree problem
Distributed minimum spanning tree problem Juho-Kustaa Kangas 24th November 2012 Abstract Given a connected weighted undirected graph, the minimum spanning tree problem asks for a spanning subtree with
More informationData Caching under Number Constraint
1 Data Caching under Number Constraint Himanshu Gupta and Bin Tang Abstract Caching can significantly improve the efficiency of information access in networks by reducing the access latency and bandwidth
More informationOn minimum m-connected k-dominating set problem in unit disc graphs
J Comb Optim (2008) 16: 99 106 DOI 10.1007/s10878-007-9124-y On minimum m-connected k-dominating set problem in unit disc graphs Weiping Shang Frances Yao Pengjun Wan Xiaodong Hu Published online: 5 December
More informationOn Performance Evaluation of Reliable Topology Control Algorithms in Mobile Ad Hoc Networks (Invited Paper)
On Performance Evaluation of Reliable Topology Control Algorithms in Mobile Ad Hoc Networks (Invited Paper) Ngo Duc Thuan 1,, Hiroki Nishiyama 1, Nirwan Ansari 2,andNeiKato 1 1 Graduate School of Information
More informationNotes for Lecture 24
U.C. Berkeley CS170: Intro to CS Theory Handout N24 Professor Luca Trevisan December 4, 2001 Notes for Lecture 24 1 Some NP-complete Numerical Problems 1.1 Subset Sum The Subset Sum problem is defined
More informationDual Power Management for Network Connectivity in Wireless Sensor Networks
Dual Power Management for Network Connectivity in Wireless Sensor Networks Yanxia Rong, Hongsik Choi and Hyeong-Ah Choi Department of Computer Science George Washington University Washington DC Department
More informationTheorem 2.9: nearest addition algorithm
There are severe limits on our ability to compute near-optimal tours It is NP-complete to decide whether a given undirected =(,)has a Hamiltonian cycle An approximation algorithm for the TSP can be used
More informationParameterized graph separation problems
Parameterized graph separation problems Dániel Marx Department of Computer Science and Information Theory, Budapest University of Technology and Economics Budapest, H-1521, Hungary, dmarx@cs.bme.hu Abstract.
More informationHEURISTIC ALGORITHMS FOR THE GENERALIZED MINIMUM SPANNING TREE PROBLEM
Proceedings of the International Conference on Theory and Applications of Mathematics and Informatics - ICTAMI 24, Thessaloniki, Greece HEURISTIC ALGORITHMS FOR THE GENERALIZED MINIMUM SPANNING TREE PROBLEM
More informationThe Full Survey on The Euclidean Steiner Tree Problem
The Full Survey on The Euclidean Steiner Tree Problem Shikun Liu Abstract The Steiner Tree Problem is a famous and long-studied problem in combinatorial optimization. However, the best heuristics algorithm
More informationChapter 5 (Week 9) The Network Layer ANDREW S. TANENBAUM COMPUTER NETWORKS FOURTH EDITION PP BLM431 Computer Networks Dr.
Chapter 5 (Week 9) The Network Layer ANDREW S. TANENBAUM COMPUTER NETWORKS FOURTH EDITION PP. 343-396 1 5.1. NETWORK LAYER DESIGN ISSUES 5.2. ROUTING ALGORITHMS 5.3. CONGESTION CONTROL ALGORITHMS 5.4.
More informationConnected Point Coverage in Wireless Sensor Networks using Robust Spanning Trees
Connected Point Coverage in Wireless Sensor Networks using Robust Spanning Trees Pouya Ostovari Department of Computer and Information Siences Temple University Philadelphia, Pennsylvania, USA Email: ostovari@temple.edu
More informationRouting. Information Networks p.1/35
Routing Routing is done by the network layer protocol to guide packets through the communication subnet to their destinations The time when routing decisions are made depends on whether we are using virtual
More informationis the Capacitated Minimum Spanning Tree
Dynamic Capacitated Minimum Spanning Trees Raja Jothi and Balaji Raghavachari Department of Computer Science, University of Texas at Dallas Richardson, TX 75083, USA raja, rbk @utdallas.edu Abstract Given
More informationECE 333: Introduction to Communication Networks Fall 2001
ECE : Introduction to Communication Networks Fall 00 Lecture : Routing and Addressing I Introduction to Routing/Addressing Lectures 9- described the main components of point-to-point networks, i.e. multiplexed
More informationSurveying Formal and Practical Approaches for Optimal Placement of Replicas on the Web
Surveying Formal and Practical Approaches for Optimal Placement of Replicas on the Web TR020701 April 2002 Erbil Yilmaz Department of Computer Science The Florida State University Tallahassee, FL 32306
More informationA Comparative study of On-Demand Data Delivery with Tables Driven and On-Demand Protocols for Mobile Ad-Hoc Network
A Comparative study of On-Demand Data Delivery with Tables Driven and On-Demand Protocols for Mobile Ad-Hoc Network Humayun Bakht Research Fellow, London School of Commerce, United Kingdom humayunbakht@yahoo.co.uk
More informationCSE 417 Branch & Bound (pt 4) Branch & Bound
CSE 417 Branch & Bound (pt 4) Branch & Bound Reminders > HW8 due today > HW9 will be posted tomorrow start early program will be slow, so debugging will be slow... Review of previous lectures > Complexity
More informationThe Effect of Neighbor Graph Connectivity on Coverage Redundancy in Wireless Sensor Networks
The Effect of Neighbor Graph Connectivity on Coverage Redundancy in Wireless Sensor Networks Eyuphan Bulut, Zijian Wang and Boleslaw K. Szymanski Department of Computer Science and Center for Pervasive
More informationCHAPTER 2 WIRELESS SENSOR NETWORKS AND NEED OF TOPOLOGY CONTROL
WIRELESS SENSOR NETWORKS AND NEED OF TOPOLOGY CONTROL 2.1 Topology Control in Wireless Sensor Networks Network topology control is about management of network topology to support network-wide requirement.
More informationApproximating Fault-Tolerant Steiner Subgraphs in Heterogeneous Wireless Networks
Approximating Fault-Tolerant Steiner Subgraphs in Heterogeneous Wireless Networks Ambreen Shahnaz and Thomas Erlebach Department of Computer Science University of Leicester University Road, Leicester LE1
More informationClustering Using Graph Connectivity
Clustering Using Graph Connectivity Patrick Williams June 3, 010 1 Introduction It is often desirable to group elements of a set into disjoint subsets, based on the similarity between the elements in the
More informationWireless Networking & Mobile Computing
Wireless Networking & Mobile Computing CS 752/852 - Spring 2012 Network Layer: Ad Hoc Routing Tamer Nadeem Dept. of Computer Science The OSI Communication Model Page 2 Spring 2012 CS 752/852 - Wireless
More informationAlgorithms for Energy-Efficient Multicasting in Static Ad Hoc Wireless Networks
Mobile Networks and Applications 6, 251 263, 2001 2001 Kluwer Academic Publishers. Manufactured in The Netherlands. Algorithms for Energy-Efficient Multicasting in Static Ad Hoc Wireless Networks JEFFREY
More informationOn the Minimum k-connectivity Repair in Wireless Sensor Networks
On the Minimum k-connectivity epair in Wireless Sensor Networks Hisham M. Almasaeid and Ahmed E. Kamal Dept. of Electrical and Computer Engineering, Iowa State University, Ames, IA 50011 Email:{hisham,kamal}@iastate.edu
More information1 The Traveling Salesperson Problem (TSP)
CS 598CSC: Approximation Algorithms Lecture date: January 23, 2009 Instructor: Chandra Chekuri Scribe: Sungjin Im In the previous lecture, we had a quick overview of several basic aspects of approximation
More informationMulticasting in the Hypercube, Chord and Binomial Graphs
Multicasting in the Hypercube, Chord and Binomial Graphs Christopher C. Cipriano and Teofilo F. Gonzalez Department of Computer Science University of California, Santa Barbara, CA, 93106 E-mail: {ccc,teo}@cs.ucsb.edu
More informationSome Applications of Graph Bandwidth to Constraint Satisfaction Problems
Some Applications of Graph Bandwidth to Constraint Satisfaction Problems Ramin Zabih Computer Science Department Stanford University Stanford, California 94305 Abstract Bandwidth is a fundamental concept
More informationApplication Layer Multicast Algorithm. 1 Introduction. Sergio Machado and Javier Ozón Universitat Politècnica de Catalunya Castelldefels.
Sergio Machado and Javier Ozón Universitat Politècnica de Catalunya Castelldefels Abstract This paper presents a multicast algorithm, called MSM-s, for point-to-multipoint transmissions. The algorithm,
More informationOn Minimizing Packet Loss Rate and Delay for Mesh-based P2P Streaming Services
On Minimizing Packet Loss Rate and Delay for Mesh-based P2P Streaming Services Zhiyong Liu, CATR Prof. Zhili Sun, UniS Dr. Dan He, UniS Denian Shi, CATR Agenda Introduction Background Problem Statement
More informationA Scalable Framework for Content Replication in Multicast-Based Content Distribution Networks
A Scalable Framework for Content Replication in Multicast-Based Content Distribution Networks Yannis Matalas 1, Nikolaos D. Dragios 2, and George T. Karetsos 2 1 Digital Media & Internet Technologies Department,
More informationAn Energy Efficient Location Service for Mobile Ad Hoc etworks
An Energ Efficient Location Service for Mobile Ad Hoc etworks Zijian Wang, Euphan Bulut and Boleslaw K. Szmanski, Department of Computer Science, Rensselaer Poltechnic Institute, Tro, NY 280 USA {wangz,
More informationA Localized Algorithm for Reducing the Size of Dominating Set in Mobile Ad Hoc Networks
A Localized Algorithm for Reducing the Size of Dominating Set in Mobile Ad Hoc Networks Yamin Li and Shietung Peng Department of Computer Science Hosei University Tokyo 18-858 Japan {yamin, speng}@k.hosei.ac.jp
More informationA Survey of Sensor Selection Schemes in Wireless Sensor Networks
A Survey of Sensor Selection Schemes in Wireless Sensor Networks Hosam Rowaihy 1, Sharanya Eswaran 1, Matthew Johnson 2, Dinesh Verma 3, Amotz Bar-Noy 2, Theodore Brown 2 and Thomas La Porta 1 1 Pennsylvania
More informationCommunication-constrained p-center Problem for Event Coverage in Theme Parks
Communication-constrained p-center Problem for Event Coverage in Theme Parks Gürkan Solmaz 1, Kemal Akkaya 2 and Damla Turgut 1 1 Department of Elect. Engineering and Computer Science, University of Central
More informationA Genetic Algorithm Applied to Graph Problems Involving Subsets of Vertices
A Genetic Algorithm Applied to Graph Problems Involving Subsets of Vertices Yaser Alkhalifah Roger L. Wainwright Department of Mathematical Department of Mathematical and Computer Sciences and Computer
More informationA Bi-Criteria Approach for Steiner s Tree Problems in Communication Networks
Workshop Investigação Operacional nas Telecomunicações Hotel Quinta das Lágrimas, 4 de Novembro, 2011Coimbra, Portugal A Bi-Criteria Approach for Steiner s Tree Problems in Communication Networks Lúcia
More informationSet Cover with Almost Consecutive Ones Property
Set Cover with Almost Consecutive Ones Property 2004; Mecke, Wagner Entry author: Michael Dom INDEX TERMS: Covering Set problem, data reduction rules, enumerative algorithm. SYNONYMS: Hitting Set PROBLEM
More informationCourse Routing Classification Properties Routing Protocols 1/39
Course 8 3. Routing Classification Properties Routing Protocols 1/39 Routing Algorithms Types Static versus dynamic Single-path versus multipath Flat versus hierarchical Host-intelligent versus router-intelligent
More informationUsing Hybrid Algorithm in Wireless Ad-Hoc Networks: Reducing the Number of Transmissions
Using Hybrid Algorithm in Wireless Ad-Hoc Networks: Reducing the Number of Transmissions R.Thamaraiselvan 1, S.Gopikrishnan 2, V.Pavithra Devi 3 PG Student, Computer Science & Engineering, Paavai College
More informationRouting with Mutual Information Accumulation in Energy-Limited Wireless Networks
Routing with Mutual Information Accumulation in Energy-Limited Wireless Networks Mahdi Shakiba-Herfeh Department of Electrical and Electronics Engineering METU, Ankara, Turkey 68 Email: mahdi@eee.metu.edu.tr
More informationBottleneck Steiner Tree with Bounded Number of Steiner Vertices
Bottleneck Steiner Tree with Bounded Number of Steiner Vertices A. Karim Abu-Affash Paz Carmi Matthew J. Katz June 18, 2011 Abstract Given a complete graph G = (V, E), where each vertex is labeled either
More informationIntroduction to Mobile Ad hoc Networks (MANETs)
Introduction to Mobile Ad hoc Networks (MANETs) 1 Overview of Ad hoc Network Communication between various devices makes it possible to provide unique and innovative services. Although this inter-device
More information7.3 Spanning trees Spanning trees [ ] 61
7.3. Spanning trees [161211-1348 ] 61 7.3 Spanning trees We know that trees are connected graphs with the minimal number of edges. Hence trees become very useful in applications where our goal is to connect
More informationAn Approximation Algorithm for Connected Dominating Set in Ad Hoc Networks
An Approximation Algorithm for Connected Dominating Set in Ad Hoc Networks Xiuzhen Cheng, Min Ding Department of Computer Science The George Washington University Washington, DC 20052, USA {cheng,minding}@gwu.edu
More informationA local area network that employs either a full mesh topology or partial mesh topology
and Ad Hoc Networks Definition A local area network that employs either a full mesh topology or partial mesh topology Full mesh topology each node is connected directly to each of the others Partial mesh
More informationRouting protocols in WSN
Routing protocols in WSN 1.1 WSN Routing Scheme Data collected by sensor nodes in a WSN is typically propagated toward a base station (gateway) that links the WSN with other networks where the data can
More informationTime-Bounded Essential Localization for Wireless Sensor Networks
IEEE TRANSACTIONS ON NETWORKING 1 Time-Bounded Essential Localization for Wireless Sensor Networks Wei Cheng, Member, IEEE, Nan Zhang, Member, IEEE, Xiuzhen Cheng, Senior Member, IEEE, Min Song, Senior
More informationTrees, Trees and More Trees
Trees, Trees and More Trees August 9, 01 Andrew B. Kahng abk@cs.ucsd.edu http://vlsicad.ucsd.edu/~abk/ How You ll See Trees in CS Trees as mathematical objects Trees as data structures Trees as tools for
More informationApproximation Algorithms
Chapter 8 Approximation Algorithms Algorithm Theory WS 2016/17 Fabian Kuhn Approximation Algorithms Optimization appears everywhere in computer science We have seen many examples, e.g.: scheduling jobs
More informationLocal Area Networks (LANs): Packets, Frames and Technologies Gail Hopkins. Part 3: Packet Switching and. Network Technologies.
Part 3: Packet Switching and Gail Hopkins Local Area Networks (LANs): Packets, Frames and Technologies Gail Hopkins Introduction Circuit Switching vs. Packet Switching LANs and shared media Star, bus and
More informationCore Membership Computation for Succinct Representations of Coalitional Games
Core Membership Computation for Succinct Representations of Coalitional Games Xi Alice Gao May 11, 2009 Abstract In this paper, I compare and contrast two formal results on the computational complexity
More informationComplexity Results on Graphs with Few Cliques
Discrete Mathematics and Theoretical Computer Science DMTCS vol. 9, 2007, 127 136 Complexity Results on Graphs with Few Cliques Bill Rosgen 1 and Lorna Stewart 2 1 Institute for Quantum Computing and School
More informationThe Impact of Clustering on the Average Path Length in Wireless Sensor Networks
The Impact of Clustering on the Average Path Length in Wireless Sensor Networks Azrina Abd Aziz Y. Ahmet Şekercioğlu Department of Electrical and Computer Systems Engineering, Monash University, Australia
More informationGeneral Algorithms for Construction of Broadcast and Multicast Trees with Applications to Wireless Networks
JOURNAL OF COMMUNICATIONS AND NETWORKS, VOL. 7, NO. 3, SEPTEMBER 2005 1 General Algorithms for Construction of Broadcast and Multicast Trees with Applications to Wireless Networks Gam D. Nguyen Abstract:
More informationThe Memetic Algorithm for The Minimum Spanning Tree Problem with Degree and Delay Constraints
The Memetic Algorithm for The Minimum Spanning Tree Problem with Degree and Delay Constraints Minying Sun*,Hua Wang* *Department of Computer Science and Technology, Shandong University, China Abstract
More informationAlgorithm Design Techniques (III)
Algorithm Design Techniques (III) Minimax. Alpha-Beta Pruning. Search Tree Strategies (backtracking revisited, branch and bound). Local Search. DSA - lecture 10 - T.U.Cluj-Napoca - M. Joldos 1 Tic-Tac-Toe
More informationarxiv: v2 [cs.dm] 3 Dec 2014
The Student/Project Allocation problem with group projects Aswhin Arulselvan, Ágnes Cseh, and Jannik Matuschke arxiv:4.035v [cs.dm] 3 Dec 04 Department of Management Science, University of Strathclyde,
More informationOn The Complexity of Virtual Topology Design for Multicasting in WDM Trees with Tap-and-Continue and Multicast-Capable Switches
On The Complexity of Virtual Topology Design for Multicasting in WDM Trees with Tap-and-Continue and Multicast-Capable Switches E. Miller R. Libeskind-Hadas D. Barnard W. Chang K. Dresner W. M. Turner
More informationDesign and Analysis of Algorithms
CSE 101, Winter 018 D/Q Greed SP s DP LP, Flow B&B, Backtrack Metaheuristics P, NP Design and Analysis of Algorithms Lecture 8: Greed Class URL: http://vlsicad.ucsd.edu/courses/cse101-w18/ Optimization
More informationTreewidth and graph minors
Treewidth and graph minors Lectures 9 and 10, December 29, 2011, January 5, 2012 We shall touch upon the theory of Graph Minors by Robertson and Seymour. This theory gives a very general condition under
More informationOptimal Backbone Generation for Robotic Relay Networks
Optimal Backbone Generation for Robotic Relay Networks Ying Zhang Palo Alto Research Center Inc. 3333 Coyote Hill Rd Palo Alto, CA 9434, USA Emails: yzhang@parc.com Mark Quilling Lockheed Martin Space
More information1. Lecture notes on bipartite matching February 4th,
1. Lecture notes on bipartite matching February 4th, 2015 6 1.1.1 Hall s Theorem Hall s theorem gives a necessary and sufficient condition for a bipartite graph to have a matching which saturates (or matches)
More informationIP Multicast Technology Overview
IP multicast is a bandwidth-conserving technology that reduces traffic by delivering a single stream of information simultaneously to potentially thousands of businesses and homes. Applications that take
More informationEmulating Web Services-Based Systems Hosted in Ad Hoc Wireless Networks
Emulating Web Services-Based Systems Hosted in Ad Hoc Wireless Networks Petr Novotny and Alexander L. Wolf Imperial College London London, UK Imperial College London Department of Computing Technical Report
More informationApproximation Algorithms
Approximation Algorithms Given an NP-hard problem, what should be done? Theory says you're unlikely to find a poly-time algorithm. Must sacrifice one of three desired features. Solve problem to optimality.
More informationMulticast Technology White Paper
Multicast Technology White Paper Keywords: Multicast, IGMP, IGMP Snooping, PIM, MBGP, MSDP, and SSM Mapping Abstract: The multicast technology implements high-efficiency point-to-multipoint data transmission
More informationEfficient Broadcast Algorithms To Reduce number of transmission Based on Probability Scheme
Efficient Broadcast s To Reduce number of transmission Based on Probability Scheme S.Tharani, R.Santhosh Abstract Two main approaches to broadcast packets in wireless ad hoc networks are static and dynamic.
More informationNetwork Coding Efficiency In The Presence Of An Intermittent Backhaul Network
IEEE ICC 2016 - Wireless Communications Symposium Network Coding Efficiency In The Presence Of An Intermittent Backhaul Network Stefan Achleitner, Thomas La Porta Computer Science and Engineering The Pennsylvania
More informationMulti-Rate Interference Sensitive and Conflict Aware Multicast in Wireless Ad hoc Networks
Multi-Rate Interference Sensitive and Conflict Aware Multicast in Wireless Ad hoc Networks Asma Ben Hassouna, Hend Koubaa, Farouk Kamoun CRISTAL Laboratory National School of Computer Science ENSI La Manouba,
More informationConstrained Minimum Spanning Tree Algorithms
December 8, 008 Introduction Graphs and MSTs revisited Minimum Spanning Tree Algorithms Algorithm of Kruskal Algorithm of Prim Constrained Minimum Spanning Trees Bounded Diameter Minimum Spanning Trees
More informationData Caching in Networks with Reading, Writing and Storage Costs
Data Caching in Networks with Reading, Writing and Storage Costs Bin Tang a, Himanshu Gupta b a Department of Electrical Engineering and Computer Science, Wichita State University, Wichita, KS 67260 b
More informationLow Power Hitch-hiking Broadcast in Ad Hoc Wireless Networks
Low Power Hitch-hiking Broadcast in Ad Hoc Wireless Networks Mihaela Cardei, Jie Wu, and Shuhui Yang Department of Computer Science and Engineering Florida Atlantic University Boca Raton, FL 33431 {mihaela,jie}@cse.fau.edu,
More informationNotes on Minimum Spanning Trees. Red Rule: Given a cycle containing no red edges, select a maximum uncolored edge on the cycle, and color it red.
COS 521 Fall 2009 Notes on Minimum Spanning Trees 1. The Generic Greedy Algorithm The generic greedy algorithm finds a minimum spanning tree (MST) by an edge-coloring process. Initially all edges are uncolored.
More informationAn Evolutionary Algorithm for the Multi-objective Shortest Path Problem
An Evolutionary Algorithm for the Multi-objective Shortest Path Problem Fangguo He Huan Qi Qiong Fan Institute of Systems Engineering, Huazhong University of Science & Technology, Wuhan 430074, P. R. China
More informationProactive data dissemination to mission sites
DOI 10.1007/s11276-012-0430-7 Proactive data dissemination to mission sites Fangfei Chen Matthew P. Johnson Amotz Bar-Noy Thomas F. La Porta Ó Springer Science+Business Media, LLC 2012 Abstract In many
More informationBenefit-based Data Caching in Ad Hoc. Networks
Benefit-based Data Caching in Ad Hoc Networks Bin Tang, Himanshu Gupta, Samir Das Computer Science Department Stony Brook University Stony Brook, NY 790 Email: {bintang,hgupta,samir}@cs.sunysb.edu Abstract
More informationG205 Fundamentals of Computer Engineering. CLASS 21, Mon. Nov Stefano Basagni Fall 2004 M-W, 1:30pm-3:10pm
G205 Fundamentals of Computer Engineering CLASS 21, Mon. Nov. 22 2004 Stefano Basagni Fall 2004 M-W, 1:30pm-3:10pm Greedy Algorithms, 1 Algorithms for Optimization Problems Sequence of steps Choices at
More informationGMNF-DVMRP: AN ENHANCED VERSION OF DISTANCE VECTOR MULTICAST ROUTING PROTOCOL
GMNF-DVMRP: AN ENHANCED VERSION OF DISTANCE VECTOR MULTICAST ROUTING PROTOCOL YUAN-CHENG LAI YING-DAR LIN AND WEI-CHE YU Department of Computer and Information Science, National Chiao Tung University,
More informationAn Efficient Approximation for the Generalized Assignment Problem
An Efficient Approximation for the Generalized Assignment Problem Reuven Cohen Liran Katzir Danny Raz Department of Computer Science Technion Haifa 32000, Israel Abstract We present a simple family of
More informationA ROUTING MECHANISM BASED ON SOCIAL NETWORKS AND BETWEENNESS CENTRALITY IN DELAY-TOLERANT NETWORKS
A ROUTING MECHANISM BASED ON SOCIAL NETWORKS AND BETWEENNESS CENTRALITY IN DELAY-TOLERANT NETWORKS ABSTRACT Zhang Huijuan and Liu Kai School of Software Engineering, Tongji University, Shanghai, China
More informationByzantine Consensus in Directed Graphs
Byzantine Consensus in Directed Graphs Lewis Tseng 1,3, and Nitin Vaidya 2,3 1 Department of Computer Science, 2 Department of Electrical and Computer Engineering, and 3 Coordinated Science Laboratory
More informationImplementation of Near Optimal Algorithm for Integrated Cellular and Ad-Hoc Multicast (ICAM)
CS230: DISTRIBUTED SYSTEMS Project Report on Implementation of Near Optimal Algorithm for Integrated Cellular and Ad-Hoc Multicast (ICAM) Prof. Nalini Venkatasubramanian Project Champion: Ngoc Do Vimal
More informationDistributed STDMA in Ad Hoc Networks
Distributed STDMA in Ad Hoc Networks Jimmi Grönkvist Swedish Defence Research Agency SE-581 11 Linköping, Sweden email: jimgro@foi.se Abstract Spatial reuse TDMA is a collision-free access scheme for ad
More informationlooking ahead to see the optimum
! Make choice based on immediate rewards rather than looking ahead to see the optimum! In many cases this is effective as the look ahead variation can require exponential time as the number of possible
More informationOn the Design of a Quality-of-Service Driven Routing Protocol for Wireless Cooperative Networks
On the Design of a Quality-of-Service Driven Routing Protocol for Wireless Cooperative Networks Zhengguo Sheng, Zhiguo Ding, and Kin K Leung Department of Electrical and Electronic Engineering Imperial
More informationAd hoc and Sensor Networks Topology control
Ad hoc and Sensor Networks Topology control Goals of this chapter Networks can be too dense too many nodes in close (radio) vicinity This chapter looks at methods to deal with such networks by Reducing/controlling
More informationReview. Some slides are in courtesy of J. Kurose and K. Ross
Review The Internet (IP) Protocol Datagram format IP fragmentation ICMP: Internet Control Message Protocol NAT: Network Address Translation Routing in the Internet Intra-AS routing: RIP and OSPF Inter-AS
More informationSolutions for the Exam 6 January 2014
Mastermath and LNMB Course: Discrete Optimization Solutions for the Exam 6 January 2014 Utrecht University, Educatorium, 13:30 16:30 The examination lasts 3 hours. Grading will be done before January 20,
More informationConstraint Satisfaction Problems
Constraint Satisfaction Problems Search and Lookahead Bernhard Nebel, Julien Hué, and Stefan Wölfl Albert-Ludwigs-Universität Freiburg June 4/6, 2012 Nebel, Hué and Wölfl (Universität Freiburg) Constraint
More informationA Network Coloring Game
A Network Coloring Game Kamalika Chaudhuri, Fan Chung 2, and Mohammad Shoaib Jamall 2 Information Theory and Applications Center, UC San Diego kamalika@soe.ucsd.edu 2 Department of Mathematics, UC San
More informationOn Distributed Algorithms for Maximizing the Network Lifetime in Wireless Sensor Networks
On Distributed Algorithms for Maximizing the Network Lifetime in Wireless Sensor Networks Akshaye Dhawan Georgia State University Atlanta, Ga 30303 akshaye@cs.gsu.edu Abstract A key challenge in Wireless
More information