Review on Energy Harvesting and Energy Management for Sustainable Wireless Sensor Networks Z.G. Wan1, Y.K. Tan2 and C. Yuen3 1
School of Electronic Information, Jiangsu University of Science and Technology, China 2 Energy Research Institute @ Nanyang Technological University (ERI@N), Singapore 3 Singapore University of Technology and Design (SUTD), Singapore E-mail:
[email protected]
Abstract - Sustainable wireless sensor networks (WSNs) are being widely used nowadays due to two key driving technologies behind them i.e. energy harvesting and energy management. Energy harvesting from environmental energy sources such as solar, wind, thermal, mechanical and so forth are introduced from the perspective of energy supply to the WSN, while energy management of WSN such as the design of MAC protocol, design of routing protocol, and dynamic power management technology are presented from the perspective of energy conservation within the WSN itself. To better understand them in details for optimizing the sustainable WSN performance, in this paper, a review of these two enabling technologies are performed. More depth research into their combined efforts for sustainable WSN is presented and then illustrated with a case study. One of the most commonly referred energy harvesting source, i.e. solar energy, and its energy management which includes a new energy forecast model of wireless sensor nodes and a new model of energy distribution in WSNs using data collection protocol is investigated and demonstrated.
I.
INTRODUCTION
In recent years, wireless sensor networks are widely used in many areas such as disaster management, infrastructure monitoring, security and surveillance, etc [1]. For these applications, the research works are mostly paying attention to the realization of functions in the designing of wireless sensor networks rather than the sustainability issue of the network. The wireless sensor node always uses power-limited battery as its energy supply. However, there are a number of nodes in the wireless sensor networks and they are always distributed in extensively wide and complex environment, it becomes very difficult to change the battery of wireless sensor nodes on deployment [2]. In order to make wireless sensor networks more practical, researchers began to study the sustainability of the wireless sensor networks, namely, try to extend the life cycle of wireless sensor networks effectively [3]. Energy harvesting and energy management are two key technologies that enable a self-sustainable wireless sensor network. There are many forms of renewable energy readily available in the environment at which the wireless sensor networks are deployed, such as solar energy, mechanical energy, thermal energy, sound energy, wind power and so on. In this paper, we conduct a review of wide varieties of energy harvesting
___________________________________ 978-1-61284-307-0/11/$26.00 ©2011 IEEE
technologies for wireless sensor networks. We mainly focus on how to transform various forms of energy existing in the environment into electrical energy that can be used to sustain the operations of the wireless sensors. Energy management technology is mainly to solve the problem of energy conservation in wireless sensor networks (WSNs) [4] [5]. Energy management usually includes optimization of medium access and routing protocols, dynamic power management etc. However, if solely relying on reducing energy consumption without energy supplement, it is very difficult to maintain long-term operation of a wireless sensor network. The objective of this paper is to explore on two key enabling technologies of a self-sustainable and self-autonomous wireless sensor network. Firstly, energy harvesting technologies for wireless sensor network, including solar, wind, sound, vibration, thermal, and electromagnetic are introduced. Secondly, energy management technology used in wireless sensor networks are summarized, which include the design of various MAC protocols, routing protocols, cross layer protocols and dynamic power management technology. Once the individual energy enhancement technology has been explored and researched, the correlation between both of these key technologies is addressed. To be able to fully optimize the WSN to be self-sustainable, rather than just energy harvesting or improved WSN energy management, it is important to further the research discussion into the combination of both energy harvesting and energy management technologies. A case study on the sustainable wireless sensor network that harvests energy from solar power, the energy model of such wireless sensor networks in green building, and the design of its data collection protocol. II. ENERGY HARVESTING TECHNOLOGY As we all know, there are many potential uses of stray energies in our living space, such as solar, wind, heat, mechanical vibration, acoustic, electromagnetic energy. These energy sources are free and pollution free. Much research work on large-scale application of environment energy including solar, wind, geothermal, etc. have already been done and the related technologies are very mature [6]. However, when the
problem changes into how to harvest and storage these natural energies in small-scale form to power miniaturized wireless sensor nodes, previous large-scale energy harvesting technologies are no longer applicable. Hence, many research works have been discussed in the literature on this energy harvesting technology for self-sustainable wireless sensor network. Some of the key progresses are described as follows. A.
Solar Energy For earth, solar energy or light energy is a kind of inexhaustible and clean energy. The basic principle of optical collection is to absorb a large number of photons by the use of photovoltaic materials. If there is enough number of photons to activate the electronic optical pool, electricity can be obtained through appropriate structural design. Because power that can be harvested is greatly depending on the light intensity, optical components are usually placed in an environment with good lighting condition in order to obtain more power. Optical components can be connected in serials to generate the required voltage. As manufacturing cost of optoelectronic components is declining, the selection of solar energy as energy source for wireless sensor networks has become a reasonable technical solution. The only disadvantage of solar energy is that it is only available during day time (for outdoor environment) or office hour (for indoor environment). A battery is needed to ensure the sensors to be operated all around the clock and the efficiency can be low on cloudy days when sun exposure is very low. A number of recent solar energy harvesting prototypes [7]-[9] for sustainable wireless sensor network are presented in Figure 1.
Figure 1: Examples of solar energy harvesting system [7] - [9] B. Wind Energy Like any of the commonly available renewable energy sources, wind energy harvesting has been widely researched
for high power applications where large wind turbinegenerators (WTGs) are used for supplying power to remote loads and grid-connected applications. Although very few research works are reported in the literature on small-scale wind energy harvesting, some efforts to generate power at a very small-scale have been made recently [10]-[12], and some are presented in Figure 2. The main disadvantage regarding wind power is unreliability factor, as the strength of the wind is not constant and unpredictable, hence it does not produce the same amount of electricity all the time. In addition, since it involves moving mechanical part, it can be noisy.
Figure 2: Examples of wind energy harvesting system [10]-[12] C. Thermal Energy Research on thermoelectric technology began in 1940's, reached its peak in 1960's. And this technology was successfully used on the spacecraft. Temperature difference generator is featured with characteristics such as small, light weight, no vibration, no noise, less maintenance and can work for long hours under harsh environment. It is suitable to act as low power less than 5W and usually mounted in a variety of unmanned surveillance sensors, tiny short-range communication devices, and medical instrumentation. At present, the relevant products have been widely used. German scientists have invented a new type of battery using the temperature of human body to produce electricity, which can provide long-term "power" for portable miniature electronic devices and eliminates the trouble of charging or replacing batteries. For example, temperature difference which equal to 5 ć between human skin and clothes can be took advantage of and provide sufficient energy for a common watch. Some examples of the thermal energy harvesting systems are presented in Figure 3.
As sustainable wireless sensor networks always use energy harvesting technology, energy management considering energy harvesting is more complex than WSNs only relying on battery supply. Until now, there are many kinds of methods for energy management. They often involve node energy management, MAC protocols, routing protocols, cross-layer optimization, etc. A. Node Energy Management For a single node in sustainable wireless sensor networks, the main objective is to make itself work permanently. [17] proposed an optimal energy management policies for Figure 3: Examples of thermal energy harvesting system [13] - energy harvesting sensor nodes. Based on its study, the [15] generated energy should be stored in a buffer. The sensor node in wireless sensor networks could sense a random field and generates a packet periodically. These packets are stored in a D. Other clean energies When sound waves encounter barriers, the sound energy can queue and transmitted using the energy available at that time. be converted into electrical energy. And according to this The energy management policy is throughout optimal and also principle, scientists design and manufacture a kind of minimizes the mean delay. Pering T proposed a new concept named dynamic voltage membrane-type receiver and connected it with the resonance. scaling (DVS) [18], which means it could dynamically adjusts When noise comes into this acoustic energy converter, the electrical power will be produced. For instance, noises of jet the supply voltage of the microprocessor according to the plane which is equal to 160dB can produce power output up to system load. DVS has a perfect effect on energy conservation, but system must have ability to measure or predict the system 100kW. Obviously, there is magnetic energy everywhere on the earth. load. Sinha A [19]firstly introduced the strategy of dynamic Therefore, the magnetic energy is a good energy for us to power management (DPM) into wireless sensor networks. utilize. New engine using magnetic energy was developed by combining generator and electric motor. It can effectively use B. MAC protocol the electromagnetic energy and pure energy to drive the MAC Protocol in WSNs is the bottom part of the network machine. This engine does not require external energy and is communication protocol. It mainly set up rules for the network an independent recycling system. Hence magnetic energy is nodes in WSNs to share digital media. Compared to traditional also an alternative source of green energy. networks, energy efficiency is a primary consideration in the design of MAC protocol of WSNs network. Some reasons of energy waste in WSNs have been concluded by [20]. They are E. Comprehensive harvesting of ambient energy Wireless sensor node that depends on only a single energy idle listening, message collision, overhearing, control-packet source is not reliable, as it does not work well when the energy overhead and overemitting, etc. For different applications of wireless sensor networks, source does not exist or disappear. In order to obtain energy as much as possible, each wireless sensor node must be designed researchers have developed a variety of different MAC with an energy collection system which can collect various protocols. Contention-based MAC protocol is featured with that nodes kinds of energy from environment [16]. Of course, the use the radio channel to send data by competition. IEEE802.11 difficulties in designation of harvesting system are mainly in MAC protocol which uses Carrier Sense Multiple Access with the following areas: (1) many kinds of energy harvesting Collision Avoidance ( CSMA/CA ) is a typical MAC protocol technologies and methods are not yet mature, there are still based on the competition. Based on IEEE802.11 MAC much innovative works which should be done by the protocol, a number of researchers proposed many kinds of researchers; (2) energy harvesting system must meet the sensor network MAC protocols based on competition, for requirements of the size of the node; (3) energy harvested by example, S-MAC protocol [21], T-MAC protocol [22], Siftnodes should be stored efficiently. MAC protocol [23]. Since MAC protocols which are based on competition III. ENERGY MANAGEMENT TECHNOLOGY always bring about conflicts, researchers have proposed Energy management is definitely an important energy-saving scheduling based MAC protocol. In this protocol, the radio means. And the objective of energy management in wireless channel should be scheduled by some scheduling algorithm networks is not only reducing energy consumption but also before sending data, and the main scheduling algorithm is called time division multiple access (TDMA). Typical balancing energy among all the nodes..
protocols include DMAC protocol [24], SMACS protocol [25], DE-MAC protocol [26], EMACS protocol [27] etc. Non-collision MAC protocol can guarantee quality of realtime because completely avoiding collision in theory. Good non-collision protocol can potentially improve throughput, reduce latency. There are two problems about non-collision protocol, one is how to use multi-channel, the other is the complexity of the protocol. TRAMA[28] is a typical Noncollision protocols. The challenge in designing a MAC protocol is to make a balance between energy conversation and flexibility of the network. Apparently, it is always easier to achieve a better performance with higher energy consumption. How to achieve the energy conversation without compromise the performance of wireless sensor networks appears to be a challenging topic. C. Routing protocol Main objective of routing protocol of wireless sensor network is to establish best path between source nodes and sink nodes. At the same time, the routing protocol should also have some characteristics such as energy saving, fault tolerance and low latency. And energy saving is definitely the main principle in routing protocol design. Some typical routing protocols in WSNs are mentioned as follows. As for cluster-based routing protocol, the network is usually divided into clusters. Each cluster consists of a cluster head and a number of cluster members. Cluster heads are responsible for coordinating works of all the nodes of the same cluster, data fusion, data forwarding etc. This protocol includes LEACH protocol [29] and TEEN protocol [30]. In the data-based routing protocols, all nodes are treated equally. This protocol is simple and robust. However, it has poor scalability. In addition, data-based routing protocol has to maintain routing tables, which will take up much storage space and increase communication throughput of the network. DD protocol [31], Rumor protocol [32] and SPIN protocol [33] are typical data-based routing protocol. Next, geographic routing protocol like GPSR protocol [34] only depends on adjacent nodes, so it is almost a stateless protocol. The nodes in a wireless networks always use the shortest Euclidean distance to establish or maintain routing table, storage are avoided and it has a short delay of data transfer. Energy routing protocol is one of routing protocols for wireless sensor networks firstly proposed. The global information of the whole network is needed for routing. However, due to energy constraints in wireless sensor networks, the node can only access the topology information of local network. As a result, it is only an ideal case. Based on this theory, Shah etc. put forward the energy aware routing protocol (EAR) [35]. DEHAR (Distributed Energy Harvesting Aware Routing Algorithm)[36] is a new adaptive and distributed routing algorithm for finding energy optimized routes in a wireless sensor network with energy harvesting. The algorithm finds an
energy efficient route from each source node to a single sink node, take into account the current energy status of the network. Each kind of routing protocol has its own advantages and disadvantages, the selection of the routing protocol should depend on network environment. D. Cross-layer optimization Generally, most communication protocols of sensor networks are designed with hierarchical structure called layers which are independent of each other, and layers are relatively simple. So networks with different systems are easy to communicate. However, in order to achieve a better efficiency with optimized system performance, cross-layer optimizations is proposed to enhance the information communication among network layers, so layers can avoid being interfered each other. Many cross-layer protocol such as E-AIMRP [37] etc. have been designed creatively. IV. SUSTAINABLE WIRELESS SENSOR NETWORKS WITH DISTRIBUTION OF ENVIRONMENT ENERGY Wireless sensor networks have been employed in various kinds of environments; it generally has a common character, i.e. sensor nodes are distributed over a relatively wide area. Therefore, the energy distribution throughout the environment will have significant impacts on the energy harvesting and data collection of sustainable wireless sensor networks. In this section, we will discuss a sustainable wireless sensor networks for indoor environmental quality monitoring of a building. We employ solar energy harvesting technology, and consider the distribution of environment energy for energy management. A. Energy harvesting technology The characteristics and performances of the renewable energy sources available in outdoor environmental conditions are very different from those found in indoor industrial and commercial environments. Within the enclosed environment like offices, hospitals, factories, etc., the energy sources are generally generated by some artificial means. Table I shows a summary of the indoor and outdoor energy sources and their characteristics. Table I: Performance of energy harvesters under indoor conditions
Table I shows that the average power that can be harvested by all the artificial energy sources is 10-100 times lower than that of the outdoor ambient energy sources. As such, these weak and uncertain indoor energy sources pose significant challenges on energy harvesting from a single energy source for sustaining the operation of the wireless sensor nodes over the entire lifetime. Further investigations were carried on the solar panel to investigate its performance at different lighting conditions. Figures 4 and 5 show the indoor solar panel’s P-V and P-R curves at different lux illuminations. Both the power curves of the solar panel in relationship with the output voltage (P-V) and the load resistance (P-R) under a range of loads from short circuit to open circuit were generated at 5 different lighting conditions ranging from 380 to 1010 lux.
C. Solar Energy Prediction and Dynamic Distribution Model in Green Building In fact, sunlight in buildings always changes regularly during a day and distribution of solar energy in buildings is characterized by region. For example, for one place in a building, there is always more light at noon than at midnight. Moreover, there is always more light in places faced east than places faced west in the morning. Therefore, considering these regularities mentioned above, a kind of model for solar energy prediction was proposed
Eh ( k +1) = Eh ( k ) + sk Δt where Eh(k) is energy harvested by node at current time, Eh(k+1) is energy prediction by Eh(k) and si . The value of si can be easily got by sensor nodes. According to the same regularities, dynamic Distribution Model in Green Building could also be induced.
F = f ( X , t)
where F is the energy distribution model, X is coordinate, t is time. Obviously, F is the function of coordinate and time. D. Data collection protocol based on Solar Energy Prediction and Dynamic Distribution Model In our designed network, data collection protocols mainly include network clustering, cluster data collection, and routing. In the stage of cluster heads competition, not only the current node's residual energy but also energy prediction model is considered in algorithm. Meanwhile, in the clustering process, we also consider the model of energy distribution.
Figure.4: P-V curves of solar panel at different lux conditions
V.
Figure 5: P-R experimental curves of solar panel at different lux conditions B. Energy model of node Energy model of node in wireless sensor networks is considered as follows:
E = Eh − Et − Er − Es
In this paper, we provide a comprehensive review on some common energy harvesting technologies of wireless sensor networks, and the introduction of energy management technology. We demonstrate an example of sustainable wireless sensor networks based on solar energy which is for green building. The challenge to harvest environment energy is discussed. ACKNOWLEDGMENT This work was supported by China National Natural Science Foundation (51008143), ERI@NTU, and Singapore University Technology and Design (grant no. SUTD-ZJU/RES/02/2011). REFERENCES
where E is the whole energy of node, Eh is energy harvested by node, Et is energy consumption for sending message, Er is energy consumption for receiving message, Es is energy consumption for sensor and computing in a node.
Eh =
CONCLUSION
[1]
n
¦ s Δt
[2]
i
i =1
where si is the energy collection efficiency of one node in wireless networks, t is time interval.
Y.K. Tan and S.K. Panda (2010). Review of Energy Harvesting Technologies for Sustainable WSN, Sustainable Wireless Sensor Network, Winston Seah (Editor) and Yen Kheng Tan (Editor-in-Chief), ISBN 978-953-307-297-5, INTECH. Available from: http://www.intechopen.com/articles/show/title/review-of-energyharvesting-technologies-for-sustainable-wsn C. Alippi, G. Anastasi, M. Di Francesco, M. Roveri, “Energy management in wireless sensor networks with energy-hungry sensors”, IEEE Instrumentation & Measurement Magazine, vol.12, issue.2, pp.1623, 2009.
[3]
V. Raghunathan, S. Ganeriwal, M. Srivastava, “Emerging techniques for long lived wireless sensor networks”, IEEE Communications Magazine, vol.44, issue.4, pp.108-114, 2006. [4] C.Y. Chong, S.P. Kumar, “Sensor networks: Evolution, opportunities, and challenges”, Proceedings of the IEEE Sensor Networks and Applications, vol.91, no.8, pp.1247-1256, 2003. [5] Ian F. Akyildiz, Weilian Su, Yogesh Sankarasubramaniam, and Erdal Cayirci, “A Survey on Sensor Networks”, IEEE Communications Magazine, 2002. [6] Y.K. Tan and S.K. Panda, “Optimized Wind Energy Harvesting System Using Resistance Emulator and Active Rectifier for Wireless Sensor Nodes”, IEEE Transactions on Power Electronics, vol.26, issue.1, pp.3850, 2011. [7] X.F. Jiang, J. Polastre and D.E. Culler, “Perpetual environmentally powered sensor networks”, Fourth International Symposium on Information Processing in Sensor Networks (IPSN), pp.463-468, 2005. [8] F.I. Simjee, P.H. Chou, “Efficient charging of supercapacitors for extended lifetime of wireless sensor nodes”, IEEE Transactions on Power Electronics, vol.23, issue.3, pp.1526-1536, 2008. [9] C. Park, P.H. Chou, “AmbiMax: Autonomous Energy Harvesting Platform for Multi-Supply Wireless Sensor Nodes”, 3rd Annual IEEE Communications Society on Sensor and Ad Hoc Communications and Networks (SECON), vol.1, pp.168-177, 2006. [10] S.Priya, C.T. Chen, Darren Fye, Jeff Zahnd, “Piezoelectric Windmill: A Novel Solution to Remote Sensing”, Japanese Journal of Applied Physics, vol.44, no.3, pp.L104L107, 2005. [11] R. Myers, M. Vickers, Kim Hyeoungwoo, S. Priya, “Small scale windmil”, Applied Physics Letters, vol.90, no.5, p.54106-1-3, 2007. [12] Y.K. Tan and S.K. Panda, "Self-Autonomous Wireless Sensor Nodes with Wind Energy Harvesting for Remote Sensing of Wind-Driven Wildfire Spread", IEEE Transactions on Instrumentation and Measurement, 2011. [13] Nicholas S. Hudak and Glenn G. Amatucci, “Small-scale energy harvesting through thermoelectric, vibration, and radio frequency power conversion”, Journal of Applied Physics, vol.103, no.10, pp.101301(124), 2008. [14] V. Leonov, T. Torfs, P. Fiorini, C. Van Hoof, “Thermoelectric Converters of Human Warmth for Self-Powered Wireless Sensor Nodes”, IEEE Sensors Journal, vol.7, issue.5, pp.650-657, 2007. [15] H.A. Sodano, G.E. Simmers, R. Dereux, D.J. Inman, “Recharging batteries using energy harvested from thermal gradients”, Journal of Intelligent Material Systems and Structures, vol.18, no.1, pp.3-10, 2007. [16] Y.K. Tan and S.K. Panda, “Energy Harvesting from Hybrid Indoor Solar and Thermal Energy Sources for Enhanced Performance of Wireless Sensor Nodes”, IEEE Transactions on Industrial Electronics, 2011. [17] Vinod Sharma, Utpal Mukherji, Vinay Joseph. “Optimal Energy Management Policies for Energy Harvesting Sensor Nodes”, IEEE Transactions on Wireless Communications, vol. 9, no. 4, pp. 1326-1336, 2010 [18] PeringT, Burd T, Brodersen R. “Dynamic voltage scaling and the design of a low-power microprocessor system”. Proceedings of Power Driven Micro architecture Workshop, attached to ISCA 98. Barcelona, Spain, 1998: 107-112. [19] Sinha A, Chandrakasan A. “Dynamic power management in wire-less sensor networks”. IEEE Design & Test of Computers, 2001, 18(2): 62-74. [20] ZHENG Guo-Qiang, LI Jian-Dong, ZHOU Zhi-Li. “Overview of MAC Protocols in Wireless Sensor Networks”. ACTA AUMATICA SINICA, 34(3): 305-316, 2008 [21] YeWˈHeidemann JˈEstrin D. “An energy-effieient MAC Protocol for wireless sensor networks”. In: Proceedings of the 21st Annual Joint Conference of the IEEE Computer and Communications Societies. New YorkˈUSA: IEEEˈ2002. 1567-1576 [22] Dam T VˈLangendoen K. “An adaptive energy-effieient MAC Protocol for wireless sensor networks”. In: Proceedings of the 1st International Conference on Embedded Networked Sensor Systems. CaliforniaˈUSA: ACMˈ2003.171-180 [23] Jamieson K, Balakrishnan H, Tay Y. Sift: “A Mac Protocol for EventDriver Wireless Sensor Network”. R Report MIT-LCS-TR-894. MIT Laboratory for Computer Scienceˈ2003 [24] Lu G,KrishnamachariB,Raghavendra C. “An adaptive enery-efficient and low-latencyMAC for data gathering in wireless sensor networks”. Santa
Fe,NewMexico. In: Proc 18th Int’lParallel and Distributed Processing Symp(IPDPS’04), 2004. 224-230. [25] SohrabiK,Gao J,AilawadhiV, “Protocols for self-organization of a wireless sensor networks”. IEEE PersonalCommunication Magazine, 2000, 7(5): 16-27. [26] KalidindiR, Ray L, Kannan1 R, “Distributed energy aware MAC layer protocol forwireless sensor networks”. LasVegas, Nevada, USA: International Conference on Wireless Networks 2003 (ICWN 03), 2003: 282-286. [27] Hoeselvan L FW.Havinga P JM. “A TDMA-based MAC protocol for WSNs”. New York, USA: Proceedings of the 2nd International Conference on Embedded Networked Sensor Systems,2004. 303-304. [28] Rajendran V, Obraczka K, Garcia-Luna-Aceves J. J. “Energy-efficient collision-free medium access control for wireless sensor networks”. LosAngeles,USA: Proc 18th Int’l Parallel and Distributed Sensor Systems (SenSys’03), 2003, 181-192. [29] Heinzelman W R, Chandrakasan A, Balakrishnan H. “Energy-efficient communication protocol for wireless microsensor networks”, Proceedings of the 33rd Annual Hawaii International Conference on System Sciences.: IEEE Computer Society, 2000: 1-10. [30] Manjeshwar A, Agrawal DP. TEEN: “A protocol for enhanced efficiency in wireless sensor networks”[C]. In:Int’l Proc.of the 15th Parallel and Distributed Processing Symp.San Francisco:IEEE Computer Society, 2001:2009-2015. [31] Intanagonwiwat C, Govindan R, Estrin D, Heidemann J. “Directed diffusion for wireless sensor networking”[C]. IEEE/ACM Trans. on Networking, 2003,11(1):2-16. [32] Braginsky D, Estrin D. “Rumor routing algorithm for sensor networks”[C].In: Proc. of the 1st workshop on sensor networks and applications. Atlanta: ACM Press, 2002:22-31. [33] Kulik J, Heinzelman WR, Balakrishnan H. “Negotiation based protocols for disseminating information in wireless sensor networks”[J].Wireless Networks,2002,8(2-3):169-185. [34] Karp B,Kung H. “GPSR: Greedy perimeter stateless routing for wireless networks”[C]. In: Proc. of the 6th Annual International Conference on Mobile Computing and Networking.Boston:ACM Press,2000: 243-254. [35] [SHAH R C, RABAEY J. “Energy aware routing for low energy Ad hoc sensor networks”[C].Proc of IEEE Wireless Communica ˉ tions and Networking Conference.2002:350-355. [36] Mikkel Koefoed Jakobsen, Jan Madsen, Michael R. Hansen. “DEHAR: a Distributed Energy Harvesting Aware Routing Algorithm for AD-hoc Multi-hop Wireless Sensor Networks”. IEEE, 2010 [37] CHEN Wei, CHENG Liang-lun, LU Xu. “Cross Layer Design Energy Efficient Protocol Integrating with MAC and Routing”. Computer Engineering, 2010, 36: 131-136
