Energy efficient and QoS based routing protocol for wireless sensor networks

1. Energy efficient and QoS based routing protocol for wireless sensor networks • By Jalel Ben-Othman, Bashir Yahya Presented by: Amrita Jyotiprada

2. Introduction Some routing techniques designed for WSNs : Negotiation based protocols – eliminate the redundant data by include high level data descriptors in the message exchange e.g. SPIN Query based - the sink node initiates the communication by broadcasting a query for data over the network QoS based - allow sensor nodes to make a tradeoff between the energy consumption and some QoS metrics before delivering the data to the sink node Multi-path based - use multiple paths rather than a single path in order to improve the network performance in terms of reliability and robustness. Multi-path routing establishes multiple paths (primarily used for load balancing, fault tolerance, bandwidth aggregation, and reduced delay) between a source-destination pair.

3. ABSTRACT The demand for real-time applications in Wireless Sensor Networks (WSNs) has made the Quality of Service (QoS) based communication protocols an important research topic. The networking protocols need to cope up with energy constraints, while providing precise QoS guarantee. Therefore, enabling real-time applications in sensor networks requires energy and QoS awareness in different layers of the protocol stack. This paper proposes an Energy Efficient and QoS aware multipath routing protocol (abbreviated as EQSR) that Maximizes the network lifetime by balancing energy consumption across multiple nodes Uses the concept of service differentiation to allow delay sensitive traffic to reach the sink node within an acceptable delay, Reduces the end to end delay by spreading out the traffic across multiple paths (supporting quality of service through multi-path routing) Increases the throughput (successful data delivery rate) through data redundancy.

4. ABSTRACT Idea contd: Recovery from node failures and achieve load balancing through splitting up the traffic across a set of available node-disjoint paths in order to efficiently balance the energy consumption over multiple sensor nodes. EQSR increases the reliability of data delivery through utilizing a light weight XOR-based FEC (Forward Error Correction) technique to provide data redundancy. Data redundancy increases resiliency to path failures and enables the protocol to recover lost data and reconstruct the original message, while avoiding any excessive delay due to data retransmissions. To predict the best next hop through the paths construction phase, EQSR uses : Residual energy Node available buffer size Signal-to-Noise Ratio (SNR) EQSR employs a queuing model that is designed to handle real-time and non-real-time traffic through service differentiation by giving real-time traffic higher priority than non-real-time traffic.

5. Multipath Routing - Benefits Load Balancing - splitting the traffic across multiple paths (helps in avoiding congestion and bottleneck problems) Reliability and fault tolerance - when primary path fails, alternative path will be used to transfer the data Highly aggregated bandwidth (multiple paths result in increased bandwidth) Minimizing end-to-end delay – achieved by dividing data into segments and sending over multiple paths.

6. Multipath Routing - Issues Nodes use shared channel in WSN and based on the MAC protocol, the neighboring node may have to wait for transmission till the channel is free. When multiple channels used, the quality of transmission may be degraded due to interference. Route Coupling – happens when two routes are located physically closed enough to interfere with each other during data communication. This may at times result in even worse performance than using single route.

7. Route Coupling – Possible solution Each node uses a directional antenna towards its target node only, then the communication in the route S-1-2-3-4-D will not affect the communication on the route S-5-6-7-8-D. When node 1 is transmitting a packet to node 2, S can transmit a packet to node 5 simultaneously. Thus the destination D will receive a packet at every time-tick with two disjoint paths using directional antenna. This decreases the overall end-to-end delay between S and D.

8. EQSR Protocol Assumptions: N identical sensor nodes. All nodes have the same transmission range, and have enough battery power to carry their computing, and communication activities. The network is fully connected and dense i.e. data can be sent from one node to another in a multi-hop basis. Each node in network is assigned a unique ID. All nodes are involved in the communication process. Sensor nodes are stationery for their lifetime. At any time, each sensor node is able to compute its residual energy (the remaining energy level), and its available buffer size. Each node is able to record the link performance between itself and the neighboring node in terms of signal-to-noise ratio (SNR).

9. EQSR Protocol Link cost function: This is used by the node to select the next hop during the path discovery phase The cost function is calculated as: where, Nx = set of neighbors of node x α, β and γ = appropriate weight factors Eresd,y = current residual energy of node y, where yϵNx Bbuffer,y = available buffer size of node y Iinterference,xy = SNR of the link between x and y Note: the residual energy of node y is considered but not x The total cost (Ctotal) for a path P consists of a set of K nodes is the sum of the individual link costs along the path. Then:

10. Path Discovery Phase The sink node starts the multiple paths discovery phase to create a set of neighbors that are able to forward data towards the sink from the source node. The constructed multi-paths are node-disjoint paths (i.e. have no common nodes except the source and the destination). Node-disjoint paths are usually preferred because they utilize the most available network resources, and hence are the most fault-tolerant. The path discovery procedure is executed according to the following phases: Initialization phase Primary Path discovery phase Alternative Paths discovery phase

11. Initialization Phase Each sensor node broadcast a HELLO message to its neighboring nodes Source Id – unique id assigned to every node. Hop Count – distance in hops for the message from its originator. Residual Energy – remaining energy of sending node. Free Buffer – available buffer size of sending node. Link Quality – expressed in terms of signal to noise ratio (SNR) for the link between any node and its neighbor. Each sensor node maintains and updates its neighbor table during this phase.

12. Primary Path Discovery Phase The sink node computes its preferred next hop node using the link cost function, and sends out a RREQ message to its most preferred next hop. In turn this preferred next hop sends a RREQ message to its most preferred next hop. This process continues until the source node is found. RREQ message structure

13. Path Discovery Phase

14. Alternative Paths Discovery Phase The sink sends alternate path RREQ message to its next most preferred neighbor. Nodes that receive more than one RREQ message, only accept the first RREQ message. Example – Node 9 sends RREQ message to node 7 but node 7 responds with an INUSE as it’s already included in the primary path. Hence, node 5 will be the next preferred neighbor

15. Path Refreshment The traffic overhead is reduced through control messages. Instead of sending KEEPALIVE message to keep multiple paths alive, the residual energy, remaining buffer size, and link quality are appended to the data message.

16. Path Selection Once paths have been constructed, some paths need to be selected from N available paths to transfer the traffic from source to destination. The number of required path (k) is calculated as: where, α = desired bound of data delivery pi (i = 1,2,....N) = probability of successful message delivery rate xα = corresponding bound from the standard normal distribution for different levels of α Fig – Some values for the bound α

17. Path Selection Then the protocol picks l paths to be used to transfer the real-time traffic and m paths for non-real-time traffic, where k = l + m Assumption: the sensor node knows the size of its traffic (both real-time and non-real-time traffic): where, Tr = size of the real-time traffic Tnr= size of the non-real-time traffic

18. Traffic Allocation & Data Transmission Protocol employs queuing model to handle both real-time and non-real-time traffic. One instant priority queue for real-time traffic and a FIFO queue for non-real-time traffic. The application layer sets the required priority level for each packet (the source node appends an extra bit to distinguish between real-time and non-real-time packets). The packet classifier directs packet into the appropriate queue. The traffic allocation scheme splits packets into equal-sized sub-packets, adds error correction codes to improve the reliability of transmission, and then transmits across the available multiple paths.

19. Traffic Allocation & Data Transmission At the sink node, the parts are collected, reassembled and the original message is recovered. Fig – Functional diagram of the EQSR protocol

20. Data Transfer Across Multiple Paths Data transfer is done through two steps: Data packet segmentation and encoding: The data packet is split up N into equal sized segments (S0, S1, S2,........., SN-1) Error correction code (C0, C1, C2,C3........., CM) are added to original message (where M < N) The data segments and error correction codes are of the same size (l bytes) and length should be multiple of 8. Correction codes are calculated based on XOR-based coding algorithm. The correction codes are computed as follows:

21. Data Transfer Across Multiple Paths Data packet segmentation and encoding contd.: Fig – Packet format (original message and correction codes) where, UP – current value of metrics (i.e. residual energy, remaining buffer size and SNR) ID – unique identifier assigned to each fragmented message MS – set 1 for all the segments except for the last segment (set to 0). When segment with MS value 0 is found, means last segment is received. Off – is used to solve the reassembling problem at the receiver side. It also helps in sequencing.

22. Data Transfer Across Multiple Paths Data packet segmentation and encoding contd.: Correction codes are generated for both real-time and non-real-time traffic (N = l for real-time and N = m for non-real-time) Traffic allocation - The paths that have a smaller end-to-end delay allocate more data segments. Assumption: the end-to-end delay of each path is inversely proportional to the available bandwidth. The nodes in the sensor network are homogeneous and the bandwidth-delay product is constant. Data forwarding and recovery: After computation, the segments and the error correction codes are sent out across multiple paths. Then the original N message segments can be recovered using XOR operation. The original message could be regenerated as:

23. Performance evaluation of EQSR Protocol The performance of the routing protocol is evaluated and compared with MCMP (Multi-Constraint Multi-Path) routing protocol. The simulation result shows that the proposed protocol achieves lower average delay, more energy savings, and higher packet delivery ratio than the MCMP protocol. NS-2 is used to implement and conduct a set of simulation experiments. The metrics used in the evaluation are: Energy consumption Delivery ratio Average delay

24. Performance evaluation of EQSR Protocol Impact of packets generation rate: Average end-to-end delay - the average time required to transfer a data packet from source node to the sink node The average packet delay of EQSR and MCMP protocols as the packet arrival rate increases. The result: EQSR differentiates network service by giving high real-time traffic absolute preferential treatment over low priority traffic. MCMP protocol outperforms the protocol in the case of non-real-time traffic, because of the overhead caused by the queuing model. For higher traffic rates the average delay increases because the EQSR protocol gives priority to process real-time traffic first.

25. Performance evaluation of EQSR Protocol Impact of packets generation rate contd Packet delivery ratio - the number of packets generated by the source node to the number of packets received by the sink node EQSR protocol outperforms the MCMP protocol; as in the case of path failures, EQSR uses forward error correction (FEC) technique to retrieve the original message.

26. Performance evaluation of EQSR Protocol Impact of packets generation rate contd Average energy consumption - the average of the energy consumed by the nodes participating in message transfer from source node to the sink node MCMP slightly outperforms EQSR protocol; because of the overhead induced by the queuing model and error codes computation.

27. Performance evaluation of EQSR Protocol Impact of node failure probability - In this simulation experiment, the behavior of the protocol in the presence of node failure is noted. Average delay It is observed that MCMP protocol is very sensitive to the increase in the node failure probability.

28. Performance evaluation of EQSR Protocol Impact of node failure probability contd Average delivery ration As the node failure probability increases, the delivery ratio of MCMP protocol drops significantly but EQSR protocol is slightly affected as EQSR employs an error detection scheme.

29. Performance evaluation of EQSR Protocol Impact of node failure probability contd Average energy consumption EQSR protocol outperforms the MCMP protocol as the EQSR protocol recovers path from path failures and reconstruct the original messages by the use of the FEC algorithm.

30. Performance evaluation of EQSR Protocol Impact of node failure probability contd Average energy consumption EQSR protocol outperforms the MCMP protocol as the EQSR protocol recovers path from path failures and reconstruct the original messages by the use of the FEC algorithm.

31. Related Work (Some earlier proposed routing protocol) Sequential Assignment Routing (SAR) protocol. SAR protocol is a multi-path routing protocol that makes routing decisions based on three factors: energy resources, QoS on each path and packet’s priority level. K. Akkaya and M. Younis proposed a cluster based QoS aware routing protocol that employs a queuing model to handle both real-time and non-real-time traffic. The protocol only considers the end-to-end delay. SPEED is another QoS based routing protocol that provides soft real-time end-to-end guarantees. Each sensor node maintains information about its neighbors and exploits geographic forwarding to find the paths. Message-initiated Constrained-Based Routing (MCBR) mechanism is proposed. MCBR is composed of explicit specifications of constraint-based destinations, route constraints and QoS requirements for messages, and a set of QoS aware meta-strategies. Felemban et al. proposed Multi-path and Multi-speed Routing Protocol (MMSPEED) for probabilistic QoS guarantee in WSNs. X. Huang and Y. Fang have proposed a multi-constrained QoS multi-path routing (MCMP) protocol that uses braided routes to deliver packets to the sink node according to certain QoS requirements expressed in terms of reliability and delay.

32. Conclusion EQSR protocol is an energy efficient and QoS aware multi-path routing protocol. Uses the multi-path with a Forward Error Correction (FEC) technique to recover from node failures. EQSR protocol uses the residual energy, node available buffer size, and signal-to-noise ratio to predict the next hop through the paths construction phase. EQSR protocol handles both real-time and non-real-time traffic efficiently, by employing a queuing model. The simulation results show that EQSR protocol achieves lower average delay, more energy savings, and higher delivery ration than the MCMP protocol. FUTURE WORK Analyze the performance of EQSR protocol and study the impact of the network size, path length, and buffer size on the performance metrics