Available online at www.sciencedirect.com ScienceDirect IERI Procedia 7 (2014 ) 42 48 2013 International Conference on Applied Computing, Computer Science, and Computer Engineering Throughput Improvement of Randomly Deployed Wireless Personal Area networks Rambabu A.Vatti a, A.N.Gaikwad b * a Ph.D Student,SCOE,Vadagaon,Pune,, Asst.Professor, Vishwakarma Institute of Technology, Pune, India. b Principal,Zeal Education Society s DCOER, Pune, India. Abstract Throughput is the key parameter used to estimatee the quality of service of the Wireless Personal Area Networks (WPAN). Throughput of the network is degraded mainly due to the packet loss. The Packet loss problem is more in the networks, in which the nodes are deployed randomly. There are three significant factors responsible for packet loss in wireless Personal Area networks. They are, interference of co-existed networks working in the same frequency band, collisions due to co- propose a located networks, and failure of intermediate nodes near the sink due to their over usage. In this paper, authors solution to solve the problem of packet loss due to over usage of the intermediate nodes. The authors propose a routing algorithm, based on the remaining energy at the intermediate nodes, to prevent the over usage of intermediate nodes. Remaining-energy based Adaptive Multi-hop Algorithm (RAMA), which takes the routing decision based on the remaining energy at each of the neighbouring nodes and adopts short distance multi hop communication to relay the data from source node to sink node. The algorithm is implemented on TI wireless sensor nodes, and the performance is compared with SimpliciTI protocol. The experimental results show that a 27% improvement in the throughput is achieved with the proposed algorithm RAMA. 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license 2013 Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer review under responsibility of Information Engineering Research Institute Selection and peer review under responsibility of Information Engineering Research Institute Keywords: Wireless Personal Area Networks; Throughput; random node deployment; packet loss; interference; Collisions. * Rambabu A.Vatti. Tel.: +919767962928. E-mail address: rambabu.vatti@vit.edu. 2212-6678 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer review under responsibility of Information Engineering Research Institute doi:10.1016/j.ieri.2014.08.008
Rambabu A.Vatti and A.N.Gaikwad / IERI Procedia 7 ( 2014 ) 42 48 43 1. Introduction 1.1. WPAN Standard The IEEE 802.15.4 is an emerging standard for Low Rate Wireless Personal Area Networks (WPANs). The main goal of this standard is to provide low power, cost effective, flexible and scalable Wireless Networks [1]. The networks transfer the data at 250kbps and support 65,536 nodes, the number of nodes will be increased in near future. The rapid progress in wireless technologies with low cost and great mobility ensuring that these technologies are used in many contexts[2]. Now- a-days, IEEE 802.15.4 WPANs are widely used in a large number of applications, such as healthcare monitoring, Industrial automation, smart home, remote metering applications, and the number of applications utilizing IEEE 802.15.4 is exponentially increasing [3],[4]. 1.2. The Packet Loss Problem In many applications, including battle field, surveillance, the nodes are randomly deployed. In such networks, the packet loss is due to coverage problem [5],[6]. However, because of low cost and low power consumption, many applications like home automation, many nodes deployed in the small area and forming the high density networks. The coverage problem does not exist in high density networks. The probability of packet loss in such a high density networks is due to collisions and congestion and node mortality due to over usage. Reducing packet loss is a complex task and the cross layer approach is required to be adopted. The control is required to be implemented at physical layer, Data link and higher layers also. The randomly deployed nodes forms into cluster tree topologies, in which the nodes near to the sink are over used due to convergent nature of the multi hop ad-hoc networks and will die much before the other nodes [7]. This causes the packet loss in the network. The packet loss problem can be solved to certain extent by using the efficient Routing protocol. However designing a good routing protocol for the wireless networks in which the nodes are deployed randomly is a very complex task[8]. Absence of central control and time varying nature of the topology are further increasing the complexity in designing a good routing protocol. This problem can be reduced by distributing the load among the nodes and routing the packets from source to sink via different intermediate nodes based on the energy available at respective nodes. It avoids over usage of any of the intermediate nodes. 1.3. The Cluster-Tree Network In Wireless Personal Area Networks, when the nodes are deployed randomly, the nodes are organized into a network topology called Cluster-tree topology. Fig. 1. The Cluster Tree Topology
44 Rambabu A.Vatti and A.N.Gaikwad / IERI Procedia 7 ( 2014 ) 42 48 In the example wireless network shown in the fig.1, the node n0 is the sink node. It has two neighbor nodes n1 and n2. The nodes n3, n4 are neighbor to the node n1. The node n5 is the neighbor to node n2. Similarly nodes n6 and n7 are neighbor nodes to node n4. In the above network, the nodes n1, n2 and n4, are intermediate nodes. They perform two functions: they transmit the data they have collected to the sink and they relay the data of the end nodes to the sink node. The end nodes n3, n5, n6 and n7 transmit the data they gathered to the sink, via the intermediate nodes. The rest of the paper is organized as follows: Section 2 describes the algorithm design, basic assumptions made in developing the algorithm, the proposed algorithm RAMA and its operation in brief. Section 3 describes the experimental setup and the hardware used. Section 4, the results, in which the screen shots and the observations recorded are presented. The result analysis is presented in Section 5. The paper is concluded in Section 6 with conclusion and scope for further research. 2. The Proposed Algorithm-RAMA 2.1. The algorithm design The basic assumptions used in the algorithm development are: 1. The randomly deployed nodes forms into a cluster tree topology 2.The network is converge cast network. 3. None of the nodes have any prior knowledge of any other node in the network. The concept of depth is used in this algorithm. The algorithm assigns a scalar value to each node, which is called as the depth of the node. The depths are assigned in such a way that they increase with their distance to sink node. The depth of the sink node is 0. All the nodes in the network transmit their data towards the nodes having relatively lesser depths, so that the packets are transmitted in the direction of the Sink node. 2.2. The algorithm-rama 1. Node n = nj, depth d = x, is gets ready to transmit data, makes the packet with sequence number = sq. 2. hop count h=0. 3. Transmit the data, if the node with d=0, is the neighbour (in the radio range). Else Go to step4. 4. Find the node nk having higher E (Battery Voltage) among all the neighbour nodes having d < x. Go to step 6. 5. If the intermediate node nk has used to relay the previous packet having sq= sq-1, delete the nk from the neighbour node list. Go to step 4. 6. Increment the hop count, h= h+1, 7. Transmit the data to nk. 8. Go to step3 2.3. The Algorithm operation The node with depth = x, gets ready with the data to transmit. It then checks whether the sink node ( the node having depth = 0), is in its radio range, and transmits the data to the sink node. If the sink node is not there in its radio range, it checks all its neighbour nodes having depth less than its own value (d<x). Then it identifies the node having more remaining energy among those nodes having depth d<x. Further check is done to see, whether the node selected has been used for relaying the immediate previous packet. If it is used, the node is discarded from the neighbour list and another node is selected from the remaining neighbour nodes using the same criteria. The node then relay the data to the selected intermediate node. The procedure is repeated at each intermediate node. The routes are established only when the nodes wish to transmit data. No attempt is
Rambabu A.Vatti and A.N.Gaikwad / IERI Procedia 7 ( 2014 ) 42 48 45 made to maintain the state of the network. All the transmitting nodes keep monitoring its neighbour nodes and keep the information of their depths and their current battery voltage. 3. The Experimental Setup Experiments are carried out on TI motes with MSP 430 microcontroller and CC2530 radio, which uses IEEE 802.15.4 compliant PHY in 2.4GHz ISM Band [9]. The motes are programmed using IAR Work Bench IDE[10]. One mote is programmed as the Sink node to receive the data and rest of the motes are programmed as nodes. The IEEE 802.15.4 standard offers unique network address to each node [11]. Figure.2. shows the block diagram of MSP ez430 RF 2530 motes used in this work. Antenna Antenna MSP430 Microcontroller CC2530 Radio CC2530 Radio MSP430 Microcontroller Fig.2a. The Transmitter Section Fig.2b. The Receiver Section The motes are then deployed randomly in basic Electronics laboratory at Vishwakarma Institute of Technology, Pune, India. The nodes are programmed to transmit the temperature data to the sink node. The care has been taken to operate the devices in no WiFi interfering channel. The WiFi interference is monitored using the InSSIDer WiFi tracking tool [12] and noted that the WiFi interference is present at that time and place is in WiFi channel 6 with RSSI( Received Signal Strength Indication) of -30 dbm. The -30bdm is very strong enough to suppress the WPAN signals as the Receiver sensitivity of the hardware devices used in this work is -92dbm[13]. So the motes are configured to work in channel 11 of the IEEE 802.15.4 PHY. The experiments are conducted using RAMA and the SimpliciTI protocol with different transmission powers ranging from -12 dbm to 0 dbm. The transmission of packets are recorded using PuTTY [14]. 4. Results 4.1. Observations Table 1. Observations with RAMA and SimpliciTI protocol Number of packets transmitted: 100 Transmission Power RAMA Number of Packets Received SimpliciTI -12dBm 75 45-8dBm 80 60-4dBm 90 60 0dBm 90 75 4.2. Screen shot
46 Rambabu A.Vatti and A.N.Gaikwad / IERI Procedia 7 ( 2014 ) 42 48 COM3-PuTTY 222 200 051 025.2 3.0 002 200 156 222 222 200 060 024.8 3.0 002 200 156 222 222 200 061 024.8 3.0 002 200 210 222 Fig. 3. The Screen shot Sink node _Id Source node_id Packet Seq-no Temperature in degrees Centigrade Node Battery voltage Hop Count Source node_id Intermediate node_id Sink node _Id Fig.4. The packet structure 5. Result analysis Packets Received Transmission power Vs number of packets recieved 100 50 0 12dBm 8dBm 4dBm 0dBm Npr SimpliciTI 45 60 65 75 Npr RAMA 75 80 90 90 Fig. 5. Relation between node transmission power and Number of packets received The observations in Table.1. are represented in graphical form in figure.5. From the graph, it is clear that for all the transmission powers, the number of packets received (Npr) of the proposed algorithm-rama is higher than the SimpliciTI. The figure.3 is the cropped screen shot, here the node n4 (node Id 200) is transmitting data packet to the sink node (sink node-id-222). The packet with seq.no.51 has been routed via the intermediate node n1 bearing node- Id 156. The packet reached to sink node in two hops. The packet with seq-no 60 is also routed through n1. But the packet with seq-no.61 is routed through another intermediate node n2 (node-id 210). It shows that the two consecutive packets are not using the same intermediate node. It avoids the over usage of intermediate nodes. The throughput of the network can be calculated using the formula Throughput of the network = (1)
Rambabu A.Vatti and A.N.Gaikwad / IERI Procedia 7 ( 2014 ) 42 48 47 Npr = number of Packets received successfully, TB = average back-off time, Rsd = data rate, L = Data packet length, LACK = ACK packet length, SIFS is Short Inter frame Spacing, DIFS = Data Inter frame Spacing. In the experiments conducted, packet length, data rate are fixed. The packets are transmitted without performing CCA(Clear Channel Assessment). So, all the variables in the denominator are same for both the experiments. Therefore, the only parameter on which the throughput depends on is the Npr, the number of packets received. Therefore, we can consider, Throughput (2) The throughput improvement = (1-6. Conclusion and Scope for further Research ) X 100 = (1-0.73) X 100 = 27%. Remaining energy based Adaptive Multi-hop Algorithm is developed to reduce the packet loss due to over usage of the intermediate nodes. The algorithm is implemented on TI motes with MSP 430 controller with CC2530 radio. Experiments are conducted for different random deployment scenarios by varying the transmission power of the nodes. The performance of the proposed algorithm RAMA and SimpliciTI protocol are compared. It is observed that the 27% more throughput is achieved with RAMA. The work can be further extended to extend the life time of the network as the nodes are not over used, the nodes can live for longer time. References [1] IEEE. 802.15.4., Standard 2006, Part 15.4: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low Rate Wireless Personal area Networks (LR WPANs), IEEE SA Standards Board 2006. [2] Giovanni Betta, Luigi Ferrigno, Experimental Investigation of the Electromagnetic Interference of ZigBee Transmitters on Measurement Instruments, IEEE Transactions on Instrumentation and Measurement,2008,pp.1-10. [3] Ashraf Darwish and Aboul Ella Hassanien, Wearable and Implantable Wireless sensor Network Solutions for Healthcare Monitoring, Sensors 2011, 11, pp.5561-5595. [4] Rambabu.V, Dr.A.N.Gaikwad, Bhooshan Humane, Throughput Improvement in Medical Ad-Hoc Sensor Networks: A Review, Challenges, Future scope for Research. International Journal of Electronics and communication Engineering & Technology (IJECET), volume.3,issue1.january-june(2012), pp. 23-28. [5] Bang Wang, Kee Chaing Chua, Vikram Srinivasan and Wei Wang, Information Coverage in Randomly Deployed Wireless Sensor Networks, IEEE Transactions on Wireless Communications, Vol. 6, No. 8, August 2007, pp.2994-3004. [6] C-F Huang and Y.C. Tseng, A survey of solutions to coverage problems in Wireless Sensor Networks, J. Internet Technol., Vol. 6, no.1, pp.1-8, 2005. [7] Faiza Nawaz and Shafaat ahmed Bazaz, Designing a Cost aware gradient based protocol for wireless sensor networks, employing data aggregation and clustering, International Journal of Advanced Computer science, Vol. 2, No. 10, pp. 369-375, October 2012. [8] Thomas Watteyne, Kris Pister, Dominique Barthel, Mischa Dohler, Isabelle Auge-Blum, Implementation of Gradient Routing in Wireless sensor Networks, IEEE GLOBECOM-2009. [9] www.ti.com. [10] www.iar.com
48 Rambabu A.Vatti and A.N.Gaikwad / IERI Procedia 7 ( 2014 ) 42 48 [11] Uros Pesovic, Joze Mohorko, Karl Benkic, Zarko Cucej, Effect of hidden Nodes in IEEE 802.15.4/ZigBee Wireless Sensor Networks, 17 th Telecommunications forum, TELFOR 2009, Serbia, Belgrade, Nov 24-26, 2009, pp.161-164. [12] www.inssider.en.softonoc.com. [13] Michal Varchola, Milos Drutarovsky, Zigbee based Home Automation, Acta Electrotechnica et Informatica, No.4, Vol. 7, 2007. [14] www.putty.org.