Energy Efficient QoS based Routing Protocols in Wireless Sensor Networks

Transcription

1 Energy Efficient QoS based Routing Protocols in Wireless Sensor Networks Aditya Chavan Dept. of Computer Engineering, D. J. Sanghvi College of Engineering, Mumbai University, Mumbai, India Shreyansh Kakadiya Dept. of Computer Engineering, D. J. Sanghvi College of Engineering, Mumbai University, Mumbai, India Prof. Abhijit Patil Assistant Professor Dept. of Computer Engineering, D. J. Sanghvi College of Engineering, Mumbai University, Mumbai, India ABSTRACT There has been a recent advancement in wireless sensor networks (WSN) that have led to many new routing protocols formulated and designed specifically for these networks.energy efficiency was considered as the fundamental goal by most of these routing protocols in order to augment the lifetime of the entire network. But due to the development and initiation of video and imaging sensors, additional challenges need to be faced. For productive management of the sensors and adequate access to the gathered measurements, transmission of imaging and video data requires bothqos aware and energy efficient routing protocols. In this paper, we present the various requirements for QoS and challenges faced in sensor networks and also provide a comparative analysis of various traditional QoS routing protocols. Keywords-Wireless networks, wireless sensor networks, quality of service, routing protocol. INTRODUCTION The twenty-first century has witnessed the growth of one of the most important technology known as the wireless sensor network. Wireless sensor networks attracted significant attention not only from academia but also industries across the globe in the past few decades [1]. WSN consists of a variety of low-cost, low-power and multifunctional Wireless sensor nodes that can carry out functions like sensing, wireless communication and computation. Communication within sensor nodes takes place over short distance in a wireless medium and unites for the accomplishment of a common task. These tasks range from environment monitoring, controlling industrial processes, and military surveillance. Even though the individual power of nodes in WSN may be limited, but combined power of the network is sufficient enough for the required target mission. Large number of WSN shows deployment of sensor nodes in an ad hoc fashion, meaning they are deployed without careful planning and engineering. A deployed sensor node should have the capability to autonomously organize itself into a wireless communication network. These sensor nodes must function without need of special attendance for a long time. The nodes are powered by batteries. Many a times it becomes difficult to change or recharge the battery of the nodes. Characteristics of WSN are denser levels of node deployment, computation and memory constraints, unreliability of sensor nodes and server power. These constraints in WSN present new opportunities in development and application of WSN. A significant amount of research has been done on important aspects of WSNs that include its architecture and protocol design. However, many fields such as energy conservation, locationing and supporting Quality of service are still unexplored. The major reason for this is the different nature of WSN compared to traditional networks. QoS is well known as an overused terms with different meanings and perspectives [2]. Perception and interpretation of the term QoS differs between various technical communities. Different applications of WSN generate different requirement for QoS for which tradition end-to-end QoS parameters may not be sufficient enough to describe them. In this case, to measure the delivery of sensor data in a more efficient and effective way, new QoS parameters will be required. Network designers can measure these parameters to decide the QoS architecture to be exploited to give QoS support for any given application. QoS REQUIREMENTS IN WSNs A new family member of wireless data networks is the Wireless sensor network (WSN) with certain 19

2 characteristics and requirements. A generic WSN is made of abundant sensor nodes across a location of interest. Each node can collect data about a condition such as pressure, temperature, lighting condition, noise, etc and send this data reports to sink node. Due to variety of applications in WSN, their QoS requirements also vary. This variation is possible to be analyses separately in different WSN application. Also, a common Qos support solution is unlikely to be present. To focus on QoS requirements imposed on the network by the applications, we initially separate QoS requirements that have other perspectives from those with network perspectives. As mentioned earlier different technical communities interpret QoS of WSN in different way. For instance, applications that involve detection of events and tracking the target, failure of acquisition of wrong data or information may have variety of reasons. Network management may be responsible; the area of event is not covered with sensors completely. Thus, we have the number of active sensors as one of the parameters of QoS. Additionally, above failure may be a result of a little functionality of sensors, thereby resulting in less data collection. This defines measurement errors or observation accuracy as a parameter that measures QoS. The problem may be loss of information resulting in information transfer as one of the parameter. But, an absolute separation of QoS perspective is achieved as a common application requirement like performance measure together with event detection may include all of them. The major focus should be on the network and how will it provide the QoS to applications, parameters that help mapping application requirements to network infrastructure and then measure the QoS support. Figure 1: A simple QoS model The two QoS perspectives can be depicted via a simple model [2] shown in Figure 1. In this model, the application layer is not concerned with how the network layer regulates its resources. They are only concerned with the services provide which directly affect the application quality. For the network layer, its main goal is to provideqos services while maximizing network resource consumption. To achieve this goal, the network is required to analyze the application requirements and deploy various network QoS mechanisms. 1. Application-specific QoS: In application-specific QoS, parameters like coverage, error measurement, optimum number of active sensorsand exposure are included. To sum up, the application will thrust specific requirements on the deployment of sensors, precision in measurement, number of active sensors that directly relate to application quality. 2. Network-specific QoS: While effectively utilizing the resources of a network, how does the underlying communication network delivers the QoS sensor data. As there are many applications of WSN, it is difficult to study all of them. For this reason, we group application in class such that all application having the same network come under one group. The three data delivery models are: event-driven, query-driven, and continuous delivery modes. A. Event-driven: Delay Intolerant (real-time), interactive, mission critical and non-end-to-end applications are major part of event-driven application. Success of application depends on observation of events after which application needs to take action quickly and reliably. B. Query-driven: Numerous query-driven applications in WSNs are query-specific, interactive, mission critical, delay tolerant and non-end-to-end applications. Queries are sent whenever there is a demand so as to save energy. The model is quite similar to event-driven model only difference is that data is pulled by the sink rather than pushed into sink (as in eventdriven). The application still needs the data quickly and reliably. C. Continuous: In this model, the sensors have a pre-specified rate to send their data continuously. Real-time data: Real-time is delay-constrained and have a particular bandwidth as required. Packet loss is tolerable to a certain extent. Non-real-time data: Periodic data from the sensor may be collected by sink when wanted. However, packet and delay need to be tolerated. 20

3 D. Hybrid Model: Applications may have the above mentioned data delivery model working together simultaneously. However, a mechanism to accommodate different types of QoS-constrained traffic is required. Class International Journal of Innovations & Advancement in Computer Science Table 1: Application Requirements These requirements are summarized in Table 1. We can also see that there are some differences in application requirements between WSNs and traditional networks. End-to-End Interactivity Delay tolerance Critically Event-driven No Yes No Yes Query-driven No Yes Query-specific Yes Continuous No No Yes Yes Hybrid No Yes No Yes QoS CHALLENGES IN WSNs As WSNs interact with the environment, their characteristics are different from other traditional wired data networks. WSNs inherit most of the QoS challenges from general wireless networks, but their peculiar characteristics give rise to specific challenges as follows: 1) Severe resource constraints: The constraints on resources involve energy utilization, bandwidth consumption, memory used, size of buffer, processing capability, and finite transmission power. 2) Unbalanced traffic: Traffic mainly flows to a small number of sink nodes from a large number of sensor nodes in most WSN applications. This provides a challenge for QoS routing. 3) Data redundancy: Data redundancy helpsrelax the reliability/robustness requirement of data delivery, but it unnecessarily utilizes energy. Data fusion or data aggregation is a solution to decrease redundancy in the data while maintaining its robustness. 4) Network dynamics: Use of power management or energy efficient schemes may give rise to network dynamics such as wireless link failures, node failures, node state transitions and node mobility. 5) Energy balance: In order to achieve a prolonged lifetime of WSNs, the energy load must be evenly divided among all sensor nodes so that the energy at a small set of sensor nodes will not be drained out soon. 6) Scalability: QoS support designed for WSNs should be able to work efficiently even when there are a large number of sensor nodes, i.e. QoS support should not deteriorate quickly when the number of nodes or their density is increased. 7) Multiple sinks: There may exist more than one sink nodes, which have their own different requirements on the network. QoS should be able to handle this challenge without compromising other factors. 8) Multiple traffic types: Having a heterogeneous set of sensor nodes gives raise to challenges for QoS support. 9) Packet criticality: QoS mechanisms may be required toset up a priority structure by differentiating packet importance. As a result, QoS support for the network should take into consideration few of the above mentioned challenges and overcome them in order to have a energy efficient routing mechanism. QoS ROUTING PROTOCOLS In QoS-based routing protocols, a balance is maintained between energy utilization and data delivery quality [3], [4].Specifically, the network has to fulfil certain QoS metrics such as delay, energy, bandwidth, etc. when performing data delivery. QoS routing is performed by reserving resourses in a wired communication that meet the QoS requirements for each individual connection. While many mechanisms have been formulated for routing QoS constrained data in wired connection 21

4 oriented networks, they cannot be applied directly to WSNs due to the limited and finite resources, such as bandwidth and energy that each sensor node has. Some of the traditional protocols used are: 1) Sequential Assignment Routing (SAR) Protocol: The SAR is the first routing protocols for WSNs that introduces the concept of QoS constrained data in the routing decisions performed for better energy efficiency [5]. Routing decision in SAR relies on three factors: QoS on each path,energy resources and the priority level of each packet. To avoid the failure of a single route, a multi-path approach is taken into consideration and localized path restoration schemes are used instead. The goal of SAR algorithm is to minimize and reduce the average weighted QoS metric throughout the lifetime of the network for efficiently routing data packets under QoS constraints. 2) SPEED Protocol: It is a QoS routing protocol for WSNs that provides soft real-time end-to-end guarantees, which can provide congestion avoidance when the network is congested [6]. The routing module in SPEED is called Stateless Geographic Non-Deterministic forwarding (SNFG) and works with four other modules at the network layer. SPEED protocol also maintains a desired delivery speed across sensor networks with a two-tier adaptation included for regulating packets sent to the MAC layer locally and deflecting traffic at the networking layer. Figure 2: SPEED Protocol Architecture (redrawn from [5]) It consists of the following components (figure 2): An API (Application Programming Interface). A delay-estimation scheme. A neighbor-beacon-exchange scheme. A Stateless Nondeterministic Geographic Forwarding (SNGF) algorithm. A Neighborhood Feedback Loop (NFL). Backpressure Rerouting. Last mile processing. Under heavy congestion, SPEED has slightly higher energy consumption mainly because SPEED delivers more packets to the destination than the other protocols when heavily congested. The main advantage of SPEED is that it performs better in terms of end-to-end delay and miss ratio. However, SPEED does not consider energy consumption in its routing protocol. Therefore, for more realistic understanding of SPEED s energy consumption, there is a need to compare it to a routing protocol that is energy-aware. 3) Energy-Aware QoS protocol: In this protocol [7] for wireless sensor networks, real-time traffic is generated and analyzed by imaging sensors. This protocol tries to meet end-toend delay during a connection by finding a least cost, energy efficient path. The link cost function used in the protocol captures the nodes transmission energy, error rate,energy reserve and other communication parameters. A class-based queuing model is employed to simultaneously support best effort and real time traffic. The model used allows service sharing for real-time and non-real-time traffic. This protocol finds a list of paths having least cost by using extended Dijkstra s algorithm and chooses a path from this list which meets the end-to-end connection delay requirement. Simulation results in [7] show that the proposed protocol consistently performs well with respect to QoS and energy metrics, however, it does not provide flexible adjusting of bandwidth sharing for different links. 4) Multi-Path and Multi-SPEED (MMSPEED) Protocol: The MMSPEED is developed for probabilistic QoS guarantee in WSNs. The QoS provisioning is performed in two domains [8]: Timeliness domain - This can be accomplished by guaranteeing multiple packet delivery speed options. 22

5 Reliability domain - This can support various reliability requirements by probabilistic multipath forwarding. These mechanisms for QoS provisioning, employ localized geographic packet forwarding augmented with dynamic compensation, which are realized in a localized way without global network information, which compromises for local decision inaccuracies as a packet moves towards its destination. The main benefit of MMSPEED is to provide thedesirable adaptability and scalability of large scale dynamic sensor networksthat it guarantees end-to-end requirements in a localized way. It can provide QoS differentiation in both timeliness and reliability domainsby significantly enhancing the effective scope of a sensor network in terms of number of flaws that meet both timeliness and reliabilityrequirements. 5) Multimedia Geographic Routing (MGR): In [9], a new architecture called mobile multimedia sensor network (MMSN) and a routing scheme called Mobile Multimedia Geographic Routing (MGR) are presented. In this architecture the mobile multimedia sensor node (MMN) is closely examined to improve the sensor network s capability for describing the event. The proposed protocol is designed to minimize the energy utilization and fulfil the constraints on the average end-to-end delay of particular applications in MMSNs. In this protocol, the main concern is to treat the delay guaranteeing as the ultimate aim with top priority for the QoS provisioning. Then, the protocol continues the attempts to minimize the energy consumption and to enlarge the lifetime of sensors. This encourages exploiting the energy delay tradeoffs for thedesign of this protocol. Thus, selecting the ideal location of current node s next hop is the protocol s main operation. In order to find this, MGR calculates the desired hop distance for next hop selection (Dhop), by dividing the current node to sink node distance (Dh t), with the current node to sink node desired hop count (Hh t). The simulation results in [9] show that for delay set to 0.035s, the MGR guarantees the QoS delay in the most cases. In addition to that MGR saves about 30 percent energy consumption and extends the network lifetime when compared to classical geographic routing. COMPARATIVE ANALYSIS A QoS-Based Routing Schemes Comparison is presented in Table 2. Therefore, SAR, SPEED and MMSPEED can provide energy efficient routing of data with guarantee quality of service considering that the nodes are not mobile. But, MGR can be more scalable than the other protocols as it can use mobility of the nodes and provide equal QoS to application and network specific metrics. Table 2: Comparison of QoS based routing protocols Scheme SAR SPEED Energy- MMSPEED MGR Aware QoS Advantages Low energy High High Provide QoS Low energy consumption and performance in performance differentiation in utilization & maintains end-to-end delay for real-time reliability and satisfies average multiple paths to and miss ratio traffic and timeliness delay constraints destination faster domains & computation improve capacity of WSNs Limitations Overhead of Low performance Low flexibility Unable to meet Treats delay with maintaining during heavy of bandwidth requirements in top priority tables and states at each node congestion high load network disregarding the rest Scalability limited limited limited limited Not limited Mobility No No No No Yes Route The path that The path that is The path that The path that is The path that Metric minimizes avg. stateless, has least cost stateless, minimizes the wt. QoS metric geographic and using geographic and delay non-deterministic dijkstra s Algorithm non-deterministic Robust low medium medium low low 23

6 CONCLUSION Routing protocols in wireless sensor networks is a new area of research, with a finite scope but rapidly flourishing set of research results. In this paper we present a comprehensive survey of QoS based routing techniques in WSNs to understand the limitations of one protocol over the other. But, they all have a common goal of trying to widen the lifetime of the sensor network while not compromising on delivery of data. One of the major threats in the design of routing protocols for WSNs is energy efficiency due to the sparse energy resources of sensors. The fundamental objective behind the routing protocol design is to keep the sensors alive for as long as possible, thus extending their life span. The energy utilization of the sensors is monopolized by data transmission and reception. Therefore, routing protocols designed for WSNs should be as energy efficient as possible to prolong the lifetime of individual sensors. The problem of the data delivery from the source to the destination is solved by QoS protocols. Thus, routing in WSNs should be carefully considered in order to secure the stability of connections and the energy consumption of the nodes. REFERENCE [1] D. Estrin, et al., Next Century Challenges: Scalable Coordination in Sensor Networks, Proc. of Mobocom 99, Seattle, August [2] A. Ganz, Z. Ganz, and K. Wongthavarawat, Multimedia Wireless Networks: Technologies, Standards, and QoS, Prentice Hall, Upper SaddleRiver, NJ, [3] K. Akkaya, M. Younis, Energy and QoS Routing for Wireless Sensor Networks, Cluster Computing, [4] G. Shafiullah, A. Agyei. P. Wolfs, A Survey of Energy-Efficient and QoS-Aware Routing Protocols for Wireless Sensor Networks, in Telecommunications, Automationand Industrial Electronics, [5] K. Sohrabi, J. Gao, V. Ailawadhi, G. Pottie, Protocols for Self-Organization of a Wireless Sensor Network, IEEE Pers. Commun., [6] T. He, J.Stankovic, C. Lu, T. Abdelzaher, SPEED: A Stateless Protocol for Real-Time Communication in Sensor Networks, in Proc. 23 rd International Conference on Distributed Computing Systems, Torodo, [7] K. Akkaya and M. Younis, An Energy-Aware QoS Routing Protocol for Wireless Sensor Networks, in Proceedings of the IEEE Workshop on Mobile and Wireless Networks (MWN 2003), Rhode Island, [8] E. Felemban, C. Lee, E. Ekici, MMSPEED: Multipath Multi-SPEED Protocol for QoS Guarantee of Reliability and Timeliness in Wireless Sensor Networks, IEEE Trans. Mobile Computing, [9] M. Chen, M. Guizani, M. Jo, Mobile Multimedia Sensor Networks: Architecture and Routing, in Proc. Mobility Management in thenetworks of the Future World, Shanghai,