Abstract: Wide range of applications such as disaster management, military and security have fueled the interest in sensor networks during the past few years.
Chapter #9 ENERGY EFFICIENT MAC PROTOCOLS FOR WIRELESS SENSOR NETWORKS
Mohamed Younis* and Tamer Nadeem** *
Department of Computer Science and Electrical Engineering, University of Maryland Baltimore County, Baltimore, MD 21205 ** Department of Computer Science, University of Maryland College Park, College Park, MD 20742
Abstract:
Wide range of applications such as disaster management, military and security have fueled the interest in sensor networks during the past few years. Sensors are typically capable of wireless communication and are significantly constrained in the amount of available resources such as energy, storage and computation. Such constraints make the design and operation of sensor networks considerably different from contemporary wireless networks, and necessitate the development of resource conscious protocols and management techniques. In this chapter we highlight the medium arbitration issues and the general requirements of link-layer protocols. We analyze the energy consumption model of widely used approaches and characterize the efficiency features. We further report on the current state of the research and highlight open issues.
Key words:
Wireless sensor networks, Energy efficient design, Link-layer protocols, Energy-aware communication, Medium access arbitration.
1.
INTRODUCTION
In recent years there have been major advances in the development of low power micro sensors. The emergence of such sensors has led practitioners to envision networking of a large set of sensors scattered over a wide area of interest1-5. A typical architecture of a sensor network consists of
2
Chapter #9
many sensing devices that are capable of probing the environment and reporting the collected data, using a radio, to a command center (sink)6,7. Sensor networks can serve many civil and military applications such as disaster management, combat field surveillance and security. In such applications, the sensors are usually powered using small batteries and deployed in unattended setup. Therefore, replacing sensor’s battery is not possible or not practical. Such energy constraints limit sensors’ lifetime and thus make efficient design and management of sensor networks a real challenge. The limitation of energy supply on-board the sensor nodes, has motivated a lot of the research on sensor networks. Such a research can be classified into two general categories addressing the main causes of energy consumption; signal processing and radio communication. The first category of work is dedicated to extending the life of the network through the selective engagement of a subset of the sensors in monitoring the environment8,9. Unselected sensor can switch to a low-power sleep mode. The goal of this work, which often is referred to as network organization, is to maintain enough live sensors to cover the area of interest for the longest possible duration. Sensor organization often involves signal processing techniques and geographical mapping in order to ensure a sufficient coverage using the least possible number of deployed sensors. The second category includes research on energy efficient radio communication. The network and link layers have received the most notable attention with the bulk of the work focusing on energy-awareness and minimization through clever route setup10-14. The main idea of energy-aware routing is to minimize the transmission power, which is proportional to distance squared, through the pursuance of multi-hop data forwarding so that the cumulative transmission energy is reduced compared to direct sensorsink communication. Energy-efficient link layer protocols tackle the energy wastage due to collisions among the radio transmission of nodes, keeping the receiver unnecessarily active and the excessive state changes of the radio circuit15-17. In this chapter we concentrate on the minimization of energy consumption at the link layer. The rest of this chapter is organized as follows. In section 2, we discuss the energy consumption models for radio circuitries identifying the major affecting parameters. Section 2 also includes a detailed analysis of the energy implications of widely-used MAC protocols for contemporary wireless networks. We investigate link-layer issues for wireless sensor networks in section 3 and enumerate the characteristics of ideal MAC protocols. Section 4 reports on the state of the research on energy efficient MAC protocols for wireless sensor networks. Finally, we conclude the
#9. Energy efficient MAC protocols for Wireless Sensor Networks
3
chapter in section 5 with a summary and a discussion of possible future research directions.
2.
ANALYSIS OF ENERGY CONSUMPTION AT THE LINK LAYER
Recent technological breakthroughs in ultra-high integration and lowpower electronics have enabled the development of tiny battery-operated (0.5~1.5 Ah, 1.2~3.6 V)1,18-20 sensors. Given the interest in using such sensors in unattended deployment in usually harsh environments, the replenishment of sensor batteries might be impossible and thus, sensors are energy constrained and their lifetime strongly depends on how long their batteries last. In addition to the sensing ability which is the main task, a sensor typically performs signal processing and data transmission. Measurements have shown that out of these three major sensor activities, a sensor expends maximum energy in data communication. This involves transmission, reception, and being idle. For example, Stemm and Katz’s measurements have concluded that the ratio of the power consumption of the CPU, memory and display systems on a PDA device to its wireless network interface is varied from 1:0.97 to 1:1.88 for different devices and network interfaces8,21. In addition, Dam and Langendoen have shown that the ratios of the processor and the radio power for their EYES sensor nodes varies from 1:12.5 when both in sleep mode to 1:4.76 when both in active mode20. Therefore, the sensor’s MAC protocol should manage the radio in optimal way to maintain sufficient sensor energy for the required mission. In this section we analyze the energy models of the radio circuitry and the different factors that affect the level of energy consumption in wireless communication. In addition, we study and compare the energy required by contemporary link layer protocols that are widely used in wireless data networks and voice communication infrastructure.
2.1
Radio Energy Consumption Models
Typically, a radio can operate in four distinct modes of operation: Idle, Receive, Transmit, and Sleep. While it is expected that the radio consumes the most energy in the “Transmit” and “Receive” modes, running in the “Idle” mode is also costly. In most cases, operating in “Idle” mode results in significantly high energy consumption, because the radio electronics have be turned on and continually decode radio signals, even noise, to detect the presence of an incoming packets. Different measurements have shown that the energy consumption ratio of these three modes could be as 1:1.05:1.421,
4
Chapter #9
1:1:2.722, and 1:2:2.523, respectively. It is thus desirable to completely shut down the radio rather than transiting into the “Idle” mode. However, switching a radio on and off very frequently can sometimes result in even more energy consumption than leaving the transceiver unit in “Idle” mode because of the start-up power. Moreover, as the transmission packet size gets smaller, the transition energy becomes dominant to the energy consumed during receiving and transmitting of packets18,19. Therefore, it is important to take this issue into account when designing energy-efficient MAC protocols. Radio energy consumption, Eradio, could be formulated as simple as19: Eradio = [(c Ptx) + b] T, where c is the transmission power coefficient, b is the constant power offset, Ptx is the power used in transmitting the signal, and T is the transmission time. A more complex model of Eradio that extends the formulation of Shih, et al., 18 can be expressed as: Eradio = [Ptx (Ttx + Ntx Tst) + Pout Ttx] + [Prx (Trx + Nrx Tst)] + Pidle Tidle, where Ptx/rx is the power consumed by the transmitter/receiver; Pout is the transmitter output power; Pidle is the power that the radio uses in the idle mode, Ttx/rx is the average time a transmitter/receiver is used each second (actual data transmission/reception time); Tidle is the average time per second a node is on and idle, Tst is the start-up time of the transceiver and Ntx/rx denotes the average number of times per second the transmitter/receiver is turned on. Ntx/rx mainly depends on the application’s traffic model and medium access arbitration scheme. Ttx/rx depends on the packet size, the channel data rate and the average number of packets sent/received per second. Pout depends on the distance that the signal travels to destination and the surrounding terrain. Pidle is typically very close to Prx. The power consumption when the radio is in the sleep mode is usually 1 to 4 orders of magnitude less than Pidle22,24,25. A detailed analysis of the power consumption of the radio circuit at the level of the individual function is provided by Wang, et al.19. Considering the above discussion and the energy consumption model, the major sources of energy wastage at the link layer could be enumerated as follows26-29. The first source is overhearing, meaning the node receive packets that is destined to other nodes. Second, the overhead of sending and receiving medium access control packets. The third source is collisions in which multiple packets get transmitted simultaneously magnifying the signal interference and thus mandating retransmissions. The fourth is the wireless noise in which packets get corrupted and need to be retransmitted or to increase the transmission power to overcome the noise level. The fifth cause of energy wastage is the excessive periods of being in an idle state. Finally, frequent switching between modes especially switching from sleep mode to an active mode leads to significant energy consumption.
#9. Energy efficient MAC protocols for Wireless Sensor Networks
2.2
5
Contemporary MAC Schemes for Wireless Networks
Medium access control covers two main issues: resource sharing method and multiple access arbitration. A number of MAC schemes are very popular in wireless networks. Most notable MAC schemes are, time division multiple access (TDMA), frequency division multiple access (FDMA), and code division multiple access (CDMA). In this subsection, we analyze the tradeoff between the performance and energy consumption of these schemes, setting the stage for further analysis in section 3 in the context of sensor networks. TDMA: In the TDMA scheme, time is divided into slots and each node is scheduled, typically through a central controller or a base-station, to transmit or receive during specific slot(s). In a node’s time slot, the full bandwidth of the channel is dedicated for data communication. As the transmission time is inversely proportional to the signal bandwidth, the transmission time (Ttx) in the radio energy consumption model will be minimized. Moreover, the node’s knowledge of its allocated communication time slots allows the node to transition to sleep mode during inactive slots. Therefore, the energy wastage due to overhearing and idle mode can be avoided. However, the TDMA scheme requires maintaining synchronized clocks throughout the network. Due to drifts, finite error can be introduced among nodes’ reference clocks causing collisions among transmitted data packets. Therefore, the base-station usually broadcasts periodic synchronization packets. Each node should not miss these synchronization packets and thus be activated prior to their transmission. Assuming that Tguard is the smallest time gap between two consecutive slots and δ is the highest possible clock drift between the clocks of two sensor, nodes must resynchronized at least once each Tguard/δ to avoid packet collisions. In other words, the nodes must be active at least δ/Tguard number of times every second to get resynchronized. Based on the radio energy consumption model, the number of times a node is reactivated to receive a packet should be reduces in order to keep energy consumption at minimum and thus Tguard must be maximized. However, increasing Tguard will decrease the effective bandwidth and increase communication latency. FDMA: The FDMA scheme eliminates the latency concern of time-based medium arbitration by allowing multiple nodes to communicate simultaneously. The total available bandwidth is divided into multiple channels that nodes are assigned to use. Collisions are minimized since nodes do not have to contend for the same channel. However, in a FDMAbased medium access control a lower bandwidth is available for each node and thus the transmission time Ttx gets extended, which is translated to an increase in the power consumption. On the other hand, since no
6
Chapter #9
synchronization mechanism is required in this scheme, Nrx becomes minimal leading to considerable energy savings. A hybrid TDMA/FDMA mechanism could be used to combine the advantages of both schemes and to overcome some of their shortcomings18. CDMA: The CDMA scheme enables simultaneous transmissions with minimal interference and allows a node to receive from multiple senders. In CDMA each node is assigned a unique code sequence for its transmission. A node spreads out its data over the entire channel bandwidth using its code sequence. The receiver's job is to dispread the bits and extracts only the data from the desired sender. Although CDMA allows transmissions to occupy the entire bandwidth of the channel at the same time, the special coding mechanism narrows the bandwidth for the node’s data. Therefore, as in FDMA, Ttx is extended. In addition to the complexity of the transceiver’s circuit in this schema, each node has to know the code sequences of potential senders. In a multi-hop networks, a node relays data and thus needs large memory to tabulate the codes of most if not all other nodes. CSMA: TDMA, FDMA and CDMA are contention-free schemes in which nodes are assigned to channels that are partitioned (i.e., in time, frequency, or code space) to avoid collisions. On the contrary, CSMA is a contentionbased MAC scheme30. In CSMA, each node is required to keep sensing the medium searching for a free channel for its transmission. When a node has packet to send, it transmit at the full channel bandwidth. No a priori coordination among nodes or synchronized clocks is required in this scheme. Using CSMA forces nodes to be awake for longer time and consequently boosts their energy consumption. In addition, data transmissions suffer from high collisions in dense networks. Increased collisions among nodes make the transmission delay unpredictable and can lead to high rate of packet drops. On the other hand, CSMA based medium arbitration is autonomous and does not require external control. The use of directional antennas has been shown to significantly enhance the performance of CSMA31.
3.
MAC-LAYER ISSUES FOR SENSOR NETWORKS
In this section, we analyze the technical considerations for designing efficient MAC protocols for wireless sensor networks. We first outline the design goals and then build on the analysis of energy consumption models for radio circuitries, discussed in section 2, to highlight the energy-related trade-offs and enumerate some of the popular performance metrics.
#9. Energy efficient MAC protocols for Wireless Sensor Networks
3.1
7
Design Goals for Efficient MAC Protocols
In this subsection we briefly discuss the main design goals for the MAC protocols of sensor networks. As will become clear, some of these goals may be conflicting and may force a trade-off. • Scalability: It is envisioned that most applications of unattended sensor networks will involve large number of nodes. Therefore, scalability of the employed protocols is crucial. The resources, i.e. time and bandwidth, sharing method and the arbitration strategy have to allow for fair access to the medium and to prevent excessive collisions. In addition, the potential for large set of communicating nodes would impose a restriction on the use of some MAC schemes such as CDMA. In almost all sensor networks, nodes relay other sensors’ data and can even perform data aggregation. Pursuing a pure CDMA scheme would require a sensor to store many code sequences, which may be impractical for tiny sensor devices with very limited computational resources. It is worth noting that the scalability of the link layer protocols is influenced by the network architecture and routing methodology. Hierarchical network structures can allow the employment of multiple resource sharing strategies and shape the network flow into patterns that can be exploited at the link layer. For example, grouping sensors into disjoint clusters allows designating non-overlapping frequency bands to clusters, similar to FDMA, and applying a TDMA or CSMA schemes for intra-cluster communication among sensors. In addition, the methodology for route setup can rule out some MAC schemes. For example, flooding-style data dissemination makes time-based medium arbitration strategy impractical. • Delay-predictability: A number of applications of sensor networks such as target tracking require delay-bounded delivery of data. Ensuing timeliness of data reception is typically handled at multiple layers in the communication stack. For example, special consideration at the network layer would alleviate the burden of long queuing time that affect the overall end-to-end delay32-34. However, the link-layer would play a major role through careful packet scheduling and predictable strategy for medium arbitration. The employed MAC scheme determines the schedule for packet transmission not only for the individual node but for the entire network. At the node level, a suitable packet classification and priority mechanism is the base for a service differentiation that allows delay centric handling of out-going packets. On the network level, a well defined and easily enforced strategy is needed to prevent inter-node competition for medium access from causing contention that makes the time for packet
8
Chapter #9 transmission and reception non-deterministic. For example, collision based medium arbitration mechanisms such as CSMA would not be appropriate for large and densely populated sensor networks since it is not known how many times a node will back off until it successfully transmit. On the other, reservation based approaches such as TDMA would be a good match despite their scalability problems.
• Adaptability: In most applications of sensor networks traffic density varies significantly over time and from part of the network to another. Such observation is valid for both event-triggered and query-based models of network operation1,3,6. For example in a forest monitoring setup only periodic status updates are sent in normal conditions while many sensor reports are generated in case of detecting a fire. In a typical query-based operation, sensors transmit only in response to requests and little traffic is generated otherwise. In addition, generated data can be subject to aggregation on route to the sink. Data aggregation can take the form of averaging the reported data, picking the maximum value, removal of redundant report, etc.1,10,11. In some cases the traffic pattern will not change if the aggregating nodes are fixed at the time of route setup and the transformation of the data is many-to-one, e.g. averaging multiple readings. However, data aggregation is mostly performed when applicable and thus can cause variability in the traffic flow. For example, elimination of redundant sensor readings filters out repetitions and prevents resource wastage. Needless to say that dropping the packets of unneeded data depends on the type of sensors and the detected events. The MAC scheme should adapt to such high fluctuation in traffic and should allow medium access rescheduling to efficiently handle burst high-priority traffic. • Energy-efficiency: Energy is a scarce resource for sensor networks. As explained earlier, medium access is a major consumer of sensor energy, especially for long-range transmission and when the radio receiver is kept on all the time. The output power of the radio transmitter is directly proportional to distance squared and can significantly magnify in a noisy environment. Energy-aware routing typically pursues multi-hop paths in order to optimize the transmission energy10-14. On the other hand, energy-conscious medium access control (MAC) can save transmission and reception energy by limiting the potential for collisions, minimizing the use of control messages, utilizing most of the available frequency band to shorten the transmission time, turning the radio into low power sleep mode when it is idle and finally, avoiding the excessive transitions among active and sleep states.
#9. Energy efficient MAC protocols for Wireless Sensor Networks •
3.2
9
Reliability: Reliable delivery of data is a classical design goal for all network infrastructures. Guaranteed packet delivery is ensured by the careful selection of error free links, avoidance of overloaded nodes, and the detection and the recovery from packet drops. There is usually a trade-off between the control traffic overhead and the level of reliability. For example, acknowledging each packet minimizes the recovery time and limits its scope, at the price of a high control traffic that can lower the effective link bandwidth, boost end-to-end delay and increase energy consumption. In wireless networks, packet drops are mainly caused by buffer overflow and signal interference. Avoiding buffer overflow is the responsibility of both the routing and the MAC protocols. Balancing load among available routes would reduce the potential for reaching the maximum capacity of the in-bound traffic buffer in relay nodes. Meanwhile, the employed medium arbitration scheme determines the buffer management strategy and has to ensure a service rate for the outbound flow that is high enough to stop the number of backlogged packets from exceeding the maximum buffer size. Packet drops due to signal interference can be minimized through the use of sufficiently high transmission power and the prevention of contention for medium access among nodes.
Energy Trade-offs and Metrics
Based on the described design goals for MAC protocols in wireless sensor networks, in this subsection we highlight sample conflicts among some of these goals with respect to the utilization of sensor’s energy and enumerate a list of metrics that can be used to assess the performance of MAC protocols. Energy Trade-offs: Based on the earlier discussion about the qualify attributes of MAC protocols for sensors networks, one can guess that it is difficult to find a protocol that can be very scalable, extremely energyefficient, flexible and highly adaptable, robust with reliable packet delivery and predictable with bounded delay. Therefore, it is expected that these attributes will be valued differently in the various applications of sensor networks. However, energy efficiency would probably stay among the top attributes given the constrained sensor’s energy supply in unattended deployments. The following is a sample of the trade-offs that the designers would encounter with respect to energy when picking an appropriate MAC protocol:
10
Chapter #9
• While our analysis of the design goals has indicated that contentionbased MAC protocols are disfavored due to scalability concerns, delay unpredictability, high signal interference and increased potential for energy wastage in collisions, they still fits well for ad-hoc network formation. In fact the lack of centralized coordination and the complexity of resource partitioning in many sensor network architectures make contention-based approaches one of the attractive choices. • One of the approaches for energy saving is to switch the radio circuitry to sleep in order to avoid energy wastage while staying in idle mode and to limit the number of transitions between sleep and active modes. However, to take advantage of such opportunities for energy reduction the link-layer should pursue a time-based medium sharing, e.g. TDMA, with accurate clock synchronization so that state transitions can be appropriately scheduled. An alternative is to use separate channels for data and control messages which requires two radios. While TDMA and similar approaches prevent medium contention and make the end-to-end delay deterministic in addition to their energy advantage, they are not suitable for many applications of sensor networks. Time-based medium sharing does not scale well since the problem of scheduling time slots subject to flow constraints is NP-hard. In addition, time-based approaches are often slow to adapt to changes in the traffic flow and density since control messages have to be prescheduled. • Despite the fact that CDMA is a good match for most of the design goals of the sensor networks in terms of collision avoidance, adaptability, and the support of bounded delay, the resources required for implementing CDMA can over-burden the design of sensors. For example, it is not expected that sensors would have large memory that can be designated to storing the codes of all deployed sensors and just limits the scalability of CDMA. In addition, the bit encoding of CDMA extends the transmission time of a message and thus increase energy consumption. Finally, the complexity and cost of the radio circuitry can also be an issue especially for largely miniaturized implementations and for disposable use of inexpensive sensors. The designer would have to give up some of the CDMA characteristics such as the uniqueness and the length of the employed codes, in order to allow feasible implementation on tiny sensor nodes and to maintain a conservative usage of sensor’s energy. Again the balanced emphasis of the design goals would have to be based on application requirements. MAC’s Performance Metrics: To assess and compare the performance of energy-conscious MAC protocols, the following mix of metrics has been deemed indicative by the research community:
#9. Energy efficient MAC protocols for Wireless Sensor Networks
11
¾ Ratio of energy wastage to total communication energy: This metric measures the efficiency of energy utilization. A protocol with low overhead, few collisions and little idle time would demonstrate small ratio of energy wastage. ¾ Average consumed energy-per packet: This metric is dedicated to assessing the quality of active mode operations. A protocol that has low average energy per packet typically encounters few packet retransmissions and suffers little contention-related collisions. It is worth noting that the use of this metric to capture the performance of the MAC protocol should exclude the effect of the output power which is proportional to distance and is usually influenced by route setup. ¾ Average and standard deviation of node lifetime: Since the usefulness of the sensor network depends on the availability of sufficiently large number of sensors to cover the area of interest, extending the lifetime of a sensor node is very important. This metric gives a good measure of the overall network lifetime. A slow depletion rate extends the life of the sensor battery and lengthens the duration that a sensor can function. Minimizing the standard deviation of node life makes the sensor coverage and the network connectivity more predictable. It should be noted that many other factors such as the date routing mechanism and the selection of active sensor contribute to this metric. However, an energy inefficient MAC protocol can still have a noticeable negative effect. ¾ Error rate: Packet drops are mainly caused by buffer overflow and signal interference. A MAC protocol, which observes the buffer size limitations and employ an effective packet scheduling at the node and network level, would minimize packet drops. In addition, signal interference can lead to low signal to noise ratios and hinder the recipient’s ability in decoding the transmitted packet. MAC protocols that reduce the potential for collisions among nodes would suffer fewer packet decoding errors. ¾ Network Throughput: Defined as the total number of packets received at the sink per time unit. Many factors, e.g. the setup of the network topology, affect this metric and not just the choice of the MAC protocol. However, the contribution of the MAC protocol is very significant. A high network throughput indicates a small error rate for packet transmission and a low level of contention for medium access.
12
4.
Chapter #9
STATE OF THE RESEARCH
Contemporary MAC layer protocols designed for wireless devices such as MACAW35 and IEEE 802.1136 are not suitable for sensor networks. These schemes will consume considerable amount of energy since they require the sensors to continuously probe the medium. In addition, these schemes require nodes to transmit control packets in order to avoid collisions. The control packet sizes will be comparable to the size of data packets, which are small in most sensor applications. On the other hand, Bluetooth37 uses a TDMA based scheme assuming that all slave nodes are within transmission range of the master node, and thus it cannot handle the multi-hop mode of transmission usually employed in sensor networks for saving energy. Power management of the radio has gained significant importance in sensor networks since the radio is a major consumer of sensor’s energy38,39. Several methods have been suggested to reduce the energy consumption of the RF circuitry in light of the energy consumption model described in section 2. Energy-conscious MAC protocols for sensor networks found in the literature can be broadly classified into two categories: contention and reservation based protocols. In this section we report on some of the pursued techniques in each category. In should be noted that we do not discuss the MAC protocols for sensor networks, which have paid little attention to energy while focusing on other performance metrics like timeliness40,41.
4.1
Contention Based Protocols
Contention based MAC protocols have been the main choice for distributed sensor architectures where the network infrastructure and access points are not well-defined. Most of the contention based MAC protocols proposed in the literature follow the operational model of CSMA, incorporating handshaking signals and a back-off mechanism to reduce the probability of collisions. However, the pursued techniques for energy conservation vary. Some focus on the energy wastage due to collisions suggesting clever power control to limit the level of interference, the use of different channels for data and control traffic, etc. Other published work exploit the energy saving through shortening the time a radio circuitry spends in idle mode. In this subsection we discuss the basic idea of some of these techniques. SmartNode: The SmartNode MAC protocol42 extends the IEEE 802.11 standards. Nodes in the SmartNode network attempt to derive the minimum transmission power to reach the other nodes from the power strengths of received packets. Each node in SmartNode system maintains a lookup table to store the unique identifiers of the neighbors it knows about with the
#9. Energy efficient MAC protocols for Wireless Sensor Networks
13
minimum transmission powers required to reach those neighbors. Before a source node transmits a data packet, it looks up for the destination node in the table. If the lookup fails, the source node transmits the ready-to-send (RTS) packet with maximum power level; otherwise, the value stored in the lookup table is used for setting the power level of the RTS packet. An additional field is added to the RTS packet to store the used transmission power. On overhearing the RTS packet, each neighbor node estimates the minimum transmission power required and stores such value in its lookup table. In order to maintain compatibility with 802.11, the destination node transmits unmodified clear-to-send (CTS) packet with the same power level indicated in the RTS packet. Hence, the source node is then able to determine the minimum transmission power required for the data packet to reach the destination node and store such value in its lookup table. Although SmartNode reduces the power transmission level that saves energy, it has some disadvantages. SmartNode still consumes energy for the RTS/CTS control packets transmissions. In addition, using different transmission power levels increases the collision rate in the network43. Fairness is another problem in SmartNode because data packets are transmitted with low power making it easy for them to get corrupted and dropped. It is worth noting that variant of this power control scheme were pursed in PCM43, PCMA44, DBTMA45, among others. PAMAS: The Power Aware Medium Access protocol and Signaling (PAMAS)16 is a CSMA based protocol in which the nodes that are not actively transmitting or receiving should power themselves off. The approach requires the nodes to use two separate channels for control and data. The control channel is used for handshaking and the data channel for regular traffic. Using two channels minimizes the potential for collisions. A node senses the data channel and responds to connection requests only if its neighbors are not transmitting or receiving. Senders that cannot establish a connection switch to sleep mode and retry later. The duration of a node stay in the sleep mode is determined based on the exchange of special probe messages on the control channel among nodes in close proximity. Switching nodes that are not participating in communication to a sleep mode has been shown to result in energy savings of up to 70%. However the protocol requires the nodes to sense the medium to transmit and does not eliminate collisions completely. In addition the protocol requires the nodes to have two separate channels (control and data), which will require two radios at each node increasing the cost, size and complexity of the sensor design. S-MAC: The sensor-MAC26 (or S-MAC) is a contention-based protocol proposed for energy-constrained networked devices. The S-MAC protocol essentially trades energy for throughput and latency by utilizing the sleep mode of the radio. Basically nodes are switched to the sleep mode for
14
Chapter #9
scheduled periods of time. When a node becomes active, it transmits its ID indicting its readiness to receive messages. A sender has to wait for hearing the ID of the receiver node prior to transmission. It is worth noting that a similar mechanism is pursued for Piconet46. If multiple senders were waiting, they have to contend for connection with the receiver node. Nodes share their sleep schedules in order to prevent a sender from spending excessive idle time awaiting a receiver to become active. Relative timestamps are exchanged to mitigate the effect for clock drifts. Throughput is reduced since only the active part of the frame is used for communication. Latency increases because a message-generating event may occur during sleep time and will be queued until the start of the next active part. In addition to latency, this approach sacrifices per hop fairness. T-MAC20 extends S-MAC by dynamically adjusting the duration between sleep intervals in which sensors are awake based on communication of nearby neighbors. In order to reduce number for switching, T-MAC introduced additional overhead control packets to prevent nodes from early sleep switching. However, the scalability of the S-MAC (and the T-MAC) is questionable. The scheduling of sleep time can be an over-burden in large networks and the end-to-end delay for packet delivery can be unpredictable. DPSM: The Dynamic Power Saving Mechanism47 is proposed by Jung and Vaidya to enhance the energy consumption of the IEEE 802.11 MAC protocol. One of the medium arbitration modes of the IEEE 802.11 standard is the distributed coordination function in which nodes content to establish connections. Time is divided into beacon intervals; each starts with a fixed size time window in which nodes announce any packets pending transmission to nodes. Nodes that are not transmitting or receiving can enter a low-power doze mode. The DPSM approach further increases the time a node spent in the doze state by dynamically determining the window size based on the node backlog and receiving activities in the previous interval. Assuming synchronized clocks, nodes exchange the size of their active window so that a sender does not wait unpredictably for a dozing receiver. In addition, the DPSM approach further qualify the effectiveness of switching to the doze mode by ensuring that the transition between the doze and active states is not going to drain more energy than the savings achieved in the doze mode. The main concern with DPSM is its scalability for large networks with dense deployment. In addition, one would argue that the required clock synchronization of all nodes does not fit the nature of a CSMA protocol. Adaptive transmission rate: A CSMA-based MAC protocol with an adaptive rate control mechanism is proposed by Woo and Culler for sensor networks48. In such protocol a node that has a data packet to transmit, senses the medium for a random period of time. If the medium is idle during this
#9. Energy efficient MAC protocols for Wireless Sensor Networks
15
period, a channel reservation process using simple RTS/CTS control packets is started. Otherwise, the node applies a binary exponential back-off scheme and then retries the transmission. During the back-off period, the node does not have to keep sensing the medium and it can turn off its radio to conserve energy. In addition, the protocol eliminates data ACK packets to further reduce energy consumption. To decrease the probability of collisions among nodes that detects the same event, each node upon detecting the event applies a random delay before starting the transmission process. A passive rate control mechanism is used to maintain fair bandwidth for both the originating data of a node and the route-thru traffic that passes through it. Basically, each node periodically attempts to inject a packet into the network. When that packet is successfully injected, it signals that the road still has capacity for more traffic and thus, the node can increase its transmission rate linearly. However if the injection of that packet was not successful, it signals that the road is jammed and the node decreases its rate of originating data using a multiplicative scheme. Route-thru traffic will adapt to the traffic of the original data using a similar mechanism. If a node injects lots of original traffic into the freeway, the route-thru traffic will be hindered and thus, the rate of transmitting route-thru traffic will decrease. This decrease will propagate deep down into the network which ultimately decreases the amount of aggregate route-thru traffic. While the approach limits collisions and strives to maintain fair bandwidth splitting among nodes, energy conservation is limited in scope. The issues of overhearing and minimizing time the radio is in idle mode are not addressed.
4.2
Reservation Based Protocols
Time based medium access has the potential of capturing most of the opportunities for energy optimization in sensor networks. As discussed in section 2, energy wastage due to overhearing, collision, idle mode and transitions between different states can be minimized if the medium access is shared on a time basis. In addition, time based medium arbitration can enhance delay predictability and limit packet drops due to interference and buffer overflow. However, the problem of scheduling access to the medium is NP-hard making the scalability of time based MAC scheme a major concern. Moreover, distributed time based medium arbitration typically introduces excessive overhead. In addition, maintaining clock synchrony among nodes is essential to enforce the schedule which is a non-trivial problem for the resource-constrained sensor nodes. Most of the time based MAC protocols proposed in the literature have focused on addressing these issues either using reservation requests over preset data routes or pursuing
16
Chapter #9
simplified heuristics to tackle the complexity of medium access scheduling. This subsection summarizes some of the published techniques. Clock Synchronization: To capture the advantages of a TDMA based scheme nodes should be synchronized. Typically, in sensor networks only the sensors that are performing some function are active at any point of time. Thus, in order to receive a synchronization message many sensors will have to switch on their radios. This will consume considerable amount of energy. Moreover, if the frequency of synchronization messages is large, the energy resources of the sensors will further diminish. Elson and Romer49 have argued that conventional synchronization schemes like NTP will consume a lot of energy in passive listening and hence are not suitable for sensor networks. It has been further concluded that no single scheme is suitable for sensor networks and the networks should adapt the synchronization scheme based upon the application. Though they have described the characteristics of what would be a good synchronization scheme for sensor networks, they have not explicitly proposed one. Jolly and Younis50 have proposed an approach for centralized network management setup where the sink establishes the routes and schedule the sensor’s transmission. The basic idea is to take advantage of the flow of routing traffic from the sink and include a reference clock value. Since all the sensors have to switch on to receive route and schedule update, the overhead that the sensors would have incurred in switching on to listen to synchronization messages will be eliminated. Because the frequency of route updates for some networks may be insufficient to maintain the desired level of clock synchronization, they have exploited the trade-off between the guard-time and the frequency of resynchronization. The guard time is a precautionary measure used to tolerate the difference in clock readings of communicating nodes. Increasing the guard time will require a sensor node to be activated earlier than its reception slot in order to tolerate a clock drift. Single-Sink Setup: The use of reservation requests has been explored for tackling the scalability of time-based medium arbitration51. Nodes that have data to transmit make a reservation request to a base-station, which responds with a traffic control message indicating medium access schedule. Nodes that are not included in the traffic control message can turn off their radio receivers. The nodes that have been assigned slots transmit in the order scheduled by the base-station. The base-station trades off latency with energy-efficiency. While it is better to bundle all transmissions from a node in consecutive time slots, the transmission of other nodes will be delayed. Arisha et al.,52 have pursued a non-reservation based approach that considers the routing paths. A depth and breadth first strategies were investigated for scheduling packets over a multi-hop paths. Their analysis has concluded that in the absence of buffer size constraints, breadth first can be very energy
#9. Energy efficient MAC protocols for Wireless Sensor Networks
17
efficient. While the focus on active routes limits the size of the scheduling problem, relying on a single controller and ignoring buffer size constraints raises concerns about applicability of these approaches for large networks. Multi-Cluster Scheduling: Jolly and Younis50 have proposed a comprehensive approach for energy-efficient time-base medium arbitration in sensor networks. Scalability is achieved through network clustering. Sensors are grouped around gateway nodes that are less energy constrained than sensors. Each gateway is assigned the responsibility of cluster management setting up multi-hop routes and assigning transmission/reception slots to sensors. Time slots are assigned to communicating sensors within the cluster to achieve efficient utilization of the energy resources. A breadth-first based slot assignment heuristics is suggested to conserve sensor’s energy by minimizing the number of transition between active and sleep modes and the duration in which active sensors are idle. Moreover, the slot assignment mechanism observes the buffering limitation at sensor’s node in order to prevent packet drops. The approach further boosts the effective network bandwidth by allowing the transmissions within the different clusters to overlap. To avoid intercluster interference, the slots assigned to nodes close to the cluster boundaries are further analyzed to detect potential collisions. After each gateway assign slots to sensors in its cluster, it broadcasts the transmission schedule to peer gateways that manage other clusters. A gateway is elected to perform the multi-cluster analysis in order to detect and resolve collision. Gateways can rotate this responsibility to balance the load or use other criteria. Collisions can be simply detected through the comparison of the assignment of each time-slot in the different clusters and the transmission range of the designated sensors. Resolving inter-cluster collisions can lead to modifying the medium access schedule within the individual cluster in a way that diminishes the intra-cluster MAC level efficiency. Alternative resolution can be through the extension of the TDMA frame size, which may boost the end-to-end latency. A simple heuristics has been proposed for minimizing the impact of the inter-cluster collision resolution. Basically slots assigned to routing trees within the same cluster are swapped in order to prevent the overlap between the transmissions of boundary nodes. Fig. 9-1 illustrates the proposed approach through an example.
18
Chapter #9 Tree1
11
Cluster 1
Tree2
Tree2 12,13,14
Tree3 1
Tree4 Tree5
Tree1
10
Tree6
Tree3
G3
16,17 5,6,7, 8,9 2
2 3
4
Tree7
4
1
3
Tree8 G2
5,6,7 ,8,9
Tree4
17 15
Cluster 2
Tree5
18,19
9
12
7,8,9, 10,11
G1
14,15, 16
6
13
1
13,14, 15,16, 17
12
Tree7
Cluster 1 2
G3
7,8
2
Tree8 14,15, 16,17
3 5,6,7 ,8,9
1 6
2,3
15,16,17 18,19 G1 14
4 6,7,8 5
11,12, 13
Cluster 2 10
13
13
4
1 G2
1
2 3,4,5
10
Tree6 11
3,4,5
14,15, 16,17
Cluster 1
Cluster 1
9
11,12
11,12 10
10
(a)
(b)
Figure #9-1. Assigned transmission slots are annotated next to each node. Before applying the inter-cluster collision avoidance heuristics in (a), the transmission of nodes on tree 4 interfere with nodes on trees 1 and 6 and similarly for nodes on trees 3 and 7. Swapping the order of slots allocated to trees in clusters 1and 3 eliminates the potential of collisions as shown in (b).
4.3
Hybrid Approaches
As discussed in section 2, each of CSMA, CDMA, FDMA and TDMA based schemes offers some advantages and suffers some shortcomings with respect to the requirements of MAC protocols for sensor networks. Not surprisingly some researchers have tried combining multiple of these schemes in order to better address these requirements. In this subsection, we briefly discuss a sample of such protocols. PARMAC: The Power-Aware Reservation based MAC (PARMAC)53 is an energy-aware protocol primarily designed for ad hoc networks and is applicable to sensor networks as well. The approach is actually a combination of contention and reservation based medium arbitration schemes. The network is divided into grids and each node is assumed to reach all the other nodes within its grid. Time is divided into fixed frames. Grids are assigned distinct frames. Each frame is composed of Reservation Period (RP) and Contention Free Period (CFP). In each RP, nodes within a grid cell exchange 3 messages to reserve the slots for data transmission and reception and the exchange of acknowledgements. Data is then sent in the CFP. The clocks of all nodes are assumed to be synchronized. The protocol saves energy by minimizing the idle time of the nodes and allowing the nodes to sleep during a CFP. Moreover, intra-grid control packets overhead and packet retransmissions are minimal, achieving significant energy savings. However, inter-grid contention is still possible and the efficiency of
#9. Energy efficient MAC protocols for Wireless Sensor Networks
19
this approach can significantly diminish if the application requires data exchange among nodes in different grids. DEANA: The Distributed Energy-Aware Node Activation (DEANA)54 approach exploits the node’s awareness of its neighbors in scheduling transmissions in a non-contending way. A single signaling channel is used and medium is arbitrated in a time-sharing manner. The protocol follows two cycles; a scheduled-access cycle used for data and a random-access cycle. The latter is used for neighbor discovery in case nodes are relocated or a node is added/removed from the network. The scheduled-access cycle are partitioned into time slots for node’s transmission. Each time slot is further split into control and data partitions. Nodes contend for time slots. The node that is allocated the slot will use the control potion to broadcast the receiver ID and then follow on with the data. The identified receiver will stay on and the other neighbor nodes will transition to a low-power sleep mode. Again the clocks of all nodes are assumed to be synchronized. However, the DEANA approach does not consider the state transition energy in the decision of whether to switch to the sleep mode or not. In addition, in a dense node deployment the protocol can be unpredictable since the set of neighbors will be large making it hard to expect the outcome of the contention for slots. Self-Organizing Sensor Networks: A self organization mechanism using a contention-free TDMA medium access protocol is proposed for sensor networks by Sohrabi and Pottie55. At the beginning nodes operate in random access mode for booting up the network. In this mode each node listens for other nodes on a fixed channel. When the first node is found, a two node sub-net network is formed and a new TDMA schedule is formed. The subnets continue to grow as they find new unattached nodes or merge with other sub-nets. Existing TDMA schedules are modified to accommodate the new formed sub-nets. After the boot-up period, nodes switch to a TDMA mode in which they follow their own local schedule and go through a sequence of transmission and reception bursts. In order to make the system scalable, no central entity and no global TDMA schedule are required. Therefore, a node can experience interferences from other nodes that selected the same transmission slots locally. To alleviate such problem, multiple separate channels, i.e. different frequencies (FDMA) or distinct spreading codes (CDMA), are used and each node randomly chooses one of those channels for its transmissions.
20
5.
Chapter #9
CONCLUSION AND OPEN RESEARCH ISSUES
Wireless sensor networks have been drawing increased attention in recent years. Sensors in such systems are typically disposable and expected to last until their energy drains. Therefore, energy is a very scarce resource for such sensor systems and has to be managed wisely in order to extend the life of the sensors for the duration of a particular mission. Energy conservation is generally targeted at all layers of communication stack. This chapter has been dedicated to the link layer analyzing the requirements, identifying energy-related design parameters, exploring trade-offs and reporting on advances made in the research community. Generally, an efficient MAC layer protocol for sensor networks should have the following characteristics: • The protocol should be scalable since most applications of sensor networks involve a large set of sensor nodes. • Collision among the transmissions of various nodes should be avoided. Collisions lead to packet drop and thus reduce throughput and cause energy wastage. • Energy consumed by the radio circuit in idle mode is almost equal to that consumed in active state. Consequently, idle mode of operation and transmission overhearing among sensors should be minimized. • To limit energy consumption during idle time, the sensors are typically switched to a sleep mode when not in use. However, active to sleep transitions and vice-versa consume considerable amount of energy. Therefore, an efficient protocol should minimize such transitions. • The protocol should minimize overhead. For example, control packets overhead and active sensing of the medium, typically performed by contention-based protocols, are inefficient in terms of energy consumption. • Packet drop due to limited buffer capacity should be prevented. • The protocol should adapt to changes in the network topology and all sensors should have a fair chance of transmitting. • Transmission of messages should be reliably handled while striving to maintain bounded latency. Most published work has tried to address few of these design goals, leaving lots of room for future research. For example, the analysis and discussion in this chapter have made it clear that there is no ideal strategy for resource sharing and medium access arbitration. We envision the pursuance of hybrid approaches to be the most promising design methodology. While there have been few attempts such as those summarized in section 4.3, extensive research is required to address the above efficiency requirements in a comprehensive way. Scalability issues are most notable among possible
#9. Energy efficient MAC protocols for Wireless Sensor Networks
21
future research directions given the growing ambition for very large deployments of tiny sensors and the widening scope of use in many application domains. Other avenues for research include the combined handling for QoS and energy constraints and the development of innovative radio designs that balance the desire for flexibility and robustness while maintaining acceptable circuit’s complexity. For example, little effort has been dedicated to exploring the effect of the directional antenna on the performance of sensor networks. Directional antennas can increase the spatial reusability of the medium and limit the scope of the overhearing problem.
REFERENCES 1. I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E.Cayirci, Wireless sensor networks: a survey, Computer Networks, 38(4), 393--422 (2002). 2. S. Tilak, N. B. Abu-Ghazaleh, and W. Heinzelman, A taxonomy of wireless microsensor network models, ACM Mobile Computing and Communication Review (MC2R), 6(2), 1—8, (2002). 3. G. J. Pottie and W. J. Kaiser, Wireless integrated network sensors, Communications of the ACM, 43(5), 51--58, (2000). 4. J. Rabaey, M. Ammer, J. Silva jr., D. Patel, and S. Roundy, PicoRadio supports ad hoc ultra low power wireless networking, IEEE Computer, 33(7), 42--48 (2000). 5. R. Min, M. Bhardwaj, S. Cho, E. Shih, A. Sinha, A. Wang, and A. Chandrakasan, Low Power Wireless Sensor Networks, in: Proceedings of International Conference on VLSI Design, Bangalore, India (January 2001). 6. D. Estrin, R. Govindan, J. Heidemann, and S. Kumar, Next century challenges: Scalable coordination in sensor networks, in: Proceedings of the 5th IEEE/ACM Annual Conference on Mobile Computing and Networks (MobiCOM'99), Seattle, WA (August 1999). 7. K. Sohrabi, J. Gao, V. Ailawadhi, and G.J. Pottie, Protocols for self-organization of a wireless sensor network, IEEE Personal Comm., 7(5), 16--27 (2000). 8. A. Buczak, V. Jamalabad, Self-organization of a heterogeneous sensor network by genetic algorithms, in: Intelligent Engineering Systems Through Artificial Neural Networks, edited by C.H. Dagli, et. al., Vol. 8, (ASME Press, 1998). 9. R. Burne, et. al, A self-organizing, cooperative UGS network for target tracking, in: Proceedings of SPIE Conference on Unattended Ground Sensor Technologies and Applications II, Orlando, FL (April 2000). 10. K. Akkaya and M. Younis, A Survey on Routing Protocols for Wireless Sensor Networks,” the Journal of Ad Hoc Networks (to appear). 11. C. Intanagonwiwat, R. Govindan, and D. Estrin, Directed diffusion: A scalable and robust communication paradigm for sensor networks, in: Proceedings of the 6th IEEE/ACM Annual Conference on Mobile Computing and Networks (MOBICOM'00), Boston, MA (August 2000). 12. R. Shah and J. Rabaey, Energy aware routing for low energy ad hoc sensor networks, in: Proceedings of IEEE Wireless Communications and Networking Conference (WCNC), Orlando, FL (March 2002).
22
Chapter #9
13. W. Heinzelman, A. Chandrakasan, and H. Balakrishnan, Energy-efficient communication protocols for wireless microsensor networks, in: Proceedings of the 33rd Hawaii International Conference on System Sciences (HICSS), (January 2000). 14. M. Younis, M. Youssef, and K. Arisha, Energy-aware routing in cluster-based sensor networks, in: Proceedings of the 10th IEEE/ACM Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS’02), Fort Worth, TX (October 2002). 15. G. Jolly and M. Younis, Energy efficient arbitration of medium access in sensor networks, in: Proceedings of the IASTED Conference on Wireless and Optical Communications (WOC 2003), Banff, Canada (July 2003). 16. S. Singh and C.S. Raghavendra, PAMAS: Power Aware Multi-Access protocol with Signaling for Ad Hoc Networks, ACM Computer Communications Review, 28(3), 5—26 (1998). 17. S. Xu and T. Saadawi, Does the IEEE 802.11 MAC protocol work well in multihop wireless ad hoc networks?, IEEE Communications Magazine, 39(6), 130—137 (2001). 18. E. Shih, S. Cho, N. Ickes, R. Min, A.Sinha, A. Wang, and A. Chandrakasan, Pysical layer driven protocol and algorithm design for energy-efficient wireless sensor networks, in: Proceedings of the 7thACM/IEEE Conference on Mobile Computing and Networks (MOBICOM’01), Rome, Italy (July 2001). 19. A. Wang, et al., Energy-efficient modulation and MAC for asymmetric microsensor systems, in: Proceedings of ISLPED 2001, Huntington Beach, CA (August 2001). 20. T. Dam, and K. Langendoen, An adaptive energy-efficient MAC protocol for wireless sensors networks, in: Proceedings of SenSys’03, Los Angles, California, USA (November 2003). 21. M. Stemm, and R. Katz, Measuring and reducing energy consumption of network interfaces in hand-held devices, IEICE Transactions on Communications, E80B(8):1125-1131, (1997). 22. M. Miller, and N. Vaidya, Minimizing energy consumption in sensor networks using a wakeup radio, in: Proceedings of the IEEE International Conference on Wireless Communications and Networks (WCNC’04), Atlanta, GA, (March 2004). 23. O. Kasten, Energy Consumption, http://www.inf.ethz.ch/~kasten/research/bathtub/ energy_consumption.html, Eldgenossische Technische Hochschule Zurich.
24.
MICA2 Mote Datasheet, www.xbow.com.
25. National Semiconductor Corporation, LMX3162 Single Chip Radio Transceiver, Evaluation Notes and Datasheet, (March 2000). 26. W. Ye, J. Heidemann and D. Estrin, An energy-efficient MAC protocol for wireless sensor networks, in: Proceedings of the IEEE INFOCOM, New York City, NY (June 2002). 27. C. Jones, K. Sivalingam, P. Agrawal, and J. C Chen, A survey of energy efficient network protocol, Wireless Networks, vol. 7 (2001). 28. P. Lettieri and B. Srivastava, Advances in wireless terminals (I), IEEE Personal Communications, (1999). 29. P. Havinga, G. Smit,, and M. Bos, Energy efficient wireless ATM design, in: Proceedings of wmATM’99 (June 1999). 30. J. Walrand, Communication networks, second edition, McGrawHill (1998), ISBN 0-25617404-0. 31. C. Sakr and T. Todd, Carrier-sense protocols for packet-switched smart antenna basestation, in: Proceedings of International Conference on Network Protocols (ICNP’97), (1997).
#9. Energy efficient MAC protocols for Wireless Sensor Networks
23
32. 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), Providence, RI (May 2003). 33. T. He, J. A. Stankovic, C. Lu, and T. Abdelzaher, SPEED: A stateless protocol for realtime communication in sensor networks, in: Proceedings of the International Conference on Distributed Computing Systems (ICDCS), Providence, RI (May 2003). 34. K. Akkaya and M. Younis, “Energy-aware Routing of Delay-constrained Data in Wireless Sensor Networks,” Journal of Communication Systems, special issue on QoS support and service differentiation in wireless networks, (to appear). 35. V. Bhagwan, et al., MACAW: A Media Access Protocol for Wireless LANs, in: the Proceedings of SIGCOMM Conference, (1994). 36. http://grouper.ieee.org/groups/802/11/main.html 37. Bluetooth. http://www.bluetooth.com 38. V. Raghunatham, et al., Energy-Aware Wireless Micro sensor Networks,” IEEE Signal processing magazine, (), 40-50 (2002). 39. E. Shih, et al., Energy-efficient link layer for wireless microsensor networks, in: the Proceedings of the Workshop on Very Large Scale Integration (WVLSI '01), Orlando, FL (April 2001). 40. M. Caccamo, L. Y. Zhang, L. Sha, and G. Buttazzo, An implicit prioritized access protocol for wireless sensor network, in: Proceedings of the IEEE Real-Time Systems Symposium (RTSS’02), Austin, TX, (December 2002). 41. C. Lu, B. Blum, T. Abdelzaher, J. Stankovic, and T. He, RAP: A real-time communication architecture for large-scale wireless sensor networks, in: Proceedings of the IEEE Real-Time and Embedded Technology and Applications Symposium (RTAS 2002), San Jose, CA (September 2002). 42. E. Poon, and B. Li, SmartNode: Achieving 802.11 MAC interoperability in powerefficient sensor networks with dynamic range adjustments, in: Proceedings of the IEEE International Conference on Distributed Computing Systems (ICDCS), Providence, Rhode Island, USA, (May 2003). 43. E. Jung and N. Vaidya, “A Power control MAC protocol for Ad Hoc networks, in: Proceedings of the 8thACM/IEEE Conference on Mobile Computing and Networks (MOBICOM’02), Atlanta, GA, (September 2002). 44. J. Monks, V. Bharghavan, and W. Hwu, "A power controlled multiple access protocol for wireless packet networks," in: Proceedings of IEEE Conference on Computer Communications (INFOCOM), Anchorage, AK, (April 2001). 45. S.-L. Wu, Y.-C. Tseng, and J.-P. Sheu, Intelligent medium access for mobile Ad Hoc networks with busy tones and power control, IEEE Journal on Selected Area in Communications, 18(9), 1647--57, (2000). 46. F. Bennett, et al., Piconet: Embedded mobile networking, IEEE Personal Communications Magazine, 4(5), 8–15, (1997). 47. E. Jung and N. Vaidya, “An energy efficient MAC protocol for wireless LANs, in: Proceedings of IEEE INFOCOM’02, New York (June 2002). 48. A. Woo and D. Culler, A transmission control scheme for media access in sensor networks,” in: Proceedings of 7th ACM/IEEE Annual Conference on Mobile Computing and Networks (MobiCOM01), Rome, Italy (July 2001). 49. J. Elson and K. Romer, Wireless sensor networks a new regime of time synchronization, in: Proceedings of the First Workshop on Hot Topics In Networks (HotNets-I), Princeton, NJ, (October 2002).
24
Chapter #9
50. G. Jolly and M. Younis, “An Energy Efficient, Scalable and Collision less MAC layer Protocol for Wireless Sensor Networks,” Journal of Wireless Networks and Mobile Computing, (to appear). 51. P. Havinga and G. Smit, Energy-efficient TDMA medium access control protocol scheduling, in: Proceedings of the Asian International Mobile Computing Conference (AMOC 2000), (2000). 52. K. Arisha, M. Youssef and M. Younis, Energy-aware TDMA-based MAC for sensor networks, in: Proceedings of the IEEE Workshop on Integrated Management of Power Aware Communications, Computing and Networking (IMPACCT 2002), New York City, NY (May 2002). 53. M. Adamou, I. Lee, and I. Shin, An energy efficient real-time medium access control protocol for wireless ad-hoc networks, in: WIP session of IEEE Real-time systems symposium (RTSS’01), London, UK (December 2001). 54. V. Rajendran, J.J. Garcia-Luna-Aceves, and K. Obraczka, An energy-efficient channel access scheduling for sensor networks, in: Proceedings The 5th International Symposium on Wireless Personal Multimedia Communication (WPMC), Honolulu, Hawaii, (October 2002). 55. K. Sohrabi and G.J. Pottie, Performance of a novel self-organization for wireless ad-hoc sensor networks, in Proceedings of the IEEE Vehicular Technology Conference, (1999).
