1 A New TDMA Based Sensor Network for Military Monitoring (MIL-MON) İlker Bekmezci, Fatih Alagöz energy consumption is to turn the radio off, when it is not used Abstract—Wireless sensor networks (WSN) is a new network . Fixed allocation methods, TDMA or FDMA is extremely family that enables to create smart environments. Although WSN suitable for this kind of network. There are some sensor has many application areas, military applications of WSN are networks based on TDMA such as LEACH , SMACS , very interesting. In this paper, a new TDMA based sensor network for military monitoring (MIL-MON) is proposed. MIL- two-tiered architecture , PACT . MON is developed to operate in large areas for acceptable LEACH is a self-organizing, adaptive clustering protocol lifetime periods. In order to realize MIL-MON, distributed time that uses randomization to distribute the energy load evenly scheduling mechanism, topology construction algorithm and among the sensors in the network . Although LEACH can rescheduling algorithm is proposed. Simulation results have reduce power consumption, there is a problem with the shown that MIL-MON can operate in large areas, in acceptable assumptions of LEACH. LEACH assumes that each node can lifetime and delay constraints. hear each other. So LEACH is not suitable for using in large Index Terms—sensor networks, TDMA, military monitoring. areas. SMACS is another sensor network that uses TDMA. In fact SMACS uses TDMA in addition to FDMA. After a series I. INTRODUCTION of handshaking signals, neighbor nodes can agree on a frequency and time pair to construct a link. SMACS produces n recent years, with the pace of the developing micro- I electro-mechanical systems (MEMS) technology, it has a scalable and reliable flat network. However, SMACS needs FDMA as well as TDMA, but sensor nodes are so tiny and been possible to integrate battery-operated sensor, computing limited that current sensor nodes cannot meet the requirements power and low power wireless communication components of SMACS. PACT uses Unifying Slot Assignment Protocol into one small size device. A sensor node collects data from (USAP) which is a TDMA scheduling scheme for on demand the environment continuously. The data collected by only one ad hoc networks. USAP is adopted for sensor networks in node are nearly useless, but the collaborative work of PACT. However, USAP is originally developed for ad hoc thousands of these nodes can be used to collect process and networks and PACT is not fully successful in power send the data about the environment. The network that is consumption for sensor networks. Another TDMA based composed of these wireless sensors is called wireless sensor sensor network proposal is two-tiered structural health network (WSN). The potential applications of wireless sensor monitoring wireless sensor network architecture. According to networks are highly varied. Environmental monitoring, this structure, there are some fixed cluster heads and sensor condition based maintenance, smart spaces, military, precision nodes. Sensor nodes are clustered around cluster heads. This agriculture, transportation, inventory tracking are just some of network is designed to use for monitoring buildings. It can not the sample application areas . One of the most common be used in large areas. application areas of sensor networks is military. In this paper, In this paper, a new TDMA based sensor network system, a new TDMA based military monitoring sensor network which can be used for military monitoring systems in a system (MIL-MON) is proposed. relatively large area, is presented. The coverage of the network Because of the unattended structure of the sensor nodes, the can be in the order of kilometer squares. In order to realize scarcest resource in sensor networks is power. Power MIL-MON, time synchronization, time slot assignment, consumption can be divided into three domains, as sensing, topology construction algorithms are developed. In order to communication, and data processing domains and dominant enhance the network in terms of delay and power usage, factor in energy consumption for sensor nodes is rescheduling and data indicator slot mechanisms are proposed. communication .. Not only transmission but also receiving The organization of the paper is as follows. In Section 2, is the main cause of energy waste. The easiest way to reduce preliminary works on TDMA based sensor networks is presented. In Section 3, the basic mechanisms and Ilker Bekmezci is with Bogazici University Computer Engineering enhancements for TDMA based sensor network will be Department. He is working in Turkish Air Force Academy (e-mail: proposed. In Section 4, the performance results of newly email@example.com). Fatih ALAGOZ is with Bogazici University Computer Engineering proposed algorithms will be outlined. Section 5 states the Department. (e-mail: firstname.lastname@example.org) conclusion and future work. 2 II. PRELIMINARY WORK from the sensor nodes and relay its data to central sink. TDMA is a technique to share the medium among multiple Because of these duties, cluster heads have to spend more users. Time frames are divided into time slots and each slot is power than regular nodes. The solution of this problem is to assigned to a user. In this way, when a user wants to access to change cluster heads periodically. Unfortunately, the solution the medium, it uses the medium in its own slot. TDMA has of quick power drain of cluster heads is a cause for spending many advantages for sensor networks. With the help of power. Periodic cluster head selection and reorganization of TDMA, node can know exactly when it should use its radio network for the new cluster heads may consume considerable circuit and sensor node can turn off its radio circuit to save energy. more power. Node uses transmitter or receiver only when it is Two tiered architecture for structural health monitoring is needed. This is why TDMA is very attractive for sensor another sensor network model that uses TDMA technique. network applications. There are some sensor network systems This sensor network architecture is designed for monitoring based on TDMA, such as LEACH, SMACS, PACT, two-tiered structural health of buildings. It is also based on clustered building monitoring system. approach. Sensor nodes are clustered and cluster heads are LEACH is a self-organizing, adaptive clustering protocol special nodes called local site masters (LSM). LSMs are that uses randomization to distribute the energy load evenly deployed one by one manually and the power source of LSMs among the sensors in the network . In LEACH, the nodes is the power source of the building. So, there is no power organize themselves into local clusters, around one node acting limitation on LSM. LSMs collect the data from their sensor as the local base station or cluster-head. A cluster head is nodes. LSMs construct a higher level of network and send the responsible for collecting data from its sensor nodes and collected data to a center. According to analysis of two tiered sending the data to a central sink. In this case, cluster heads sensor network, the expected lifetime of network is about 18 use more energy than the regular nodes. In order to prevent months . It is very impressive and acceptable. However, quick power drain of cluster heads, cluster heads are changed because of LSM limitations, the application of this network in periodically. The complete algorithm of LEACH is as follows: open areas is extremely difficult. There must be lots of LSM units and these LSM units should be deployed manually. While (there is at least one node with power) do Moreover, LSMs must have unlimited source power. Because Cluster_Head_Selection_and_Advertisement of these limitations, it is not suitable to use this system as Cluster_Construction military monitoring system. Create_and_Send_Time_Scheduling SMACS is a TDMA based sensor network that produces flat While (data transmission period is not exceeded) topology. In fact, SMACS uses TDMA and FDMA at the same Data_Transmission time. After deployment of sensor nodes, they wake up at Wend random times and they start to listen to a certain frequency. Wend This frequency can be called as establishment frequency. After listening to the establishment frequency for a random Comparison study shows that LEACH achieves between 7x period of time, if it can not receive any invitation signal from and 8x reduction in energy compared with direct other nodes, it transmits a short invitation packet that contains communication and between 4x and 8x reduction in energy some basic data about the node. This packet is called TYPE1 compared with MTE routing. In addition to reducing energy message. If a node can receive a TYPE1 message from other dissipation, LEACH successfully distributes energy-usage nodes and if it wants to establish a link, it responds TYPE1 among the nodes in the network such that the nodes die that it will be an invitee. This response is TYPE2 message. randomly and at essentially in the same rate . There may be more than one TYPE2 messages for a TYPE1 Although LEACH achieves power consumption decrease message. At that particular time, the node that transmits dramatically, the assumption of LEACH system is not realistic. TYPE1 message has to decide which node it should choose. According to LEACH, each node can receive signals from all After its decision, it sends a response to TYPE2 that includes the other nodes. However, it is not possible in every data about the node that is selected. This is TYPE3 message. environment, especially in larger areas. If sensor network is At the end, TYPE4 is sent by the invitee. In TYPE3 and designed to operate for relatively large areas as in MIL-MON, TYPE4 phases, nodes agree on a certain schedule to every node cannot receive all signals. communicate. This schedule is a time and frequency pair. In Another unrealistic assumption of LEACH is about the this case, the collision probability of schedules can be existence of data. LEACH assumes that every node has data to minimized. send every time. However, in most of the time, there is no data As times goes on, there will be some subnets in the wireless to send in sensor network. For example, MIL-MON is sensor networks. These subnets unite with each other, when designed for intrusion detection and the existing of an intruder new links are established. At the end, all sensor nodes is not likely to occur most of the time. construct a connected network. Another disadvantage of LEACH is the need for changing SMACS is an infrastructure building protocol that forms a cluster heads. Cluster heads are responsible for collecting data flat topology for sensor networks. It is a distributed protocol 3 which enables a collection of nodes to discover their neighbors protocols that are developed for a WSN may not be optimal and establish transmission/reception schedules, without any for another WSN . Sensor network systems should be need for master node . Although flat topology has some application specific . A sample application scenario can advantages, it has also some difficulties. Flat topology requires help to understand the assumptions and mechanisms of a separate network layer, while cluster based approach has MIL-MON. implicit network layer in itself. However, the most serious In a typical military monitoring scenario, large numbers of handicap of SMACS is that there is no sensor node that can unattended, limited powered sensor nodes are deployed near support the requirements of SMACS. According to SMACS, the borders of a base. Sensor nodes organize themselves, so node must support frequency multiplexer and there is no such that, when sensor nodes detect an intruder, they send their data a node up to now. A sensor network should be applicable on to a sink. In this way, soldiers can defend against the intruders. the existing node models. In most of the cases, sink is a PDA or a laptop. PACT (Power Aware Clustered TDMA) is a TDMA based sensor network that uses passive clustering. Its TDMA Here are the assumptions of the sensor network system: scheduling scheme is based on Unifying Slot Assignment 1) Sensor nodes will be immobile. Mobile cases can be Protocol (USAP) which is a time scheduling algorithm for on investigated for further analysis. demand ad hoc networks. PACT can be considered as a sensor 2) Power consumption model of sensor node is the same as network version of USAP in terms of time scheduling. This is described in . This model is one of the mostly used why it is not fully optimized for power consumption. In PACT, models in sensor network simulation analysis researches. every node must listen to all the control slots of a time frame. 3) Radio channel is symmetric. The number of control slot must be equal to time slot number 4) Sink node’s power source theoretically infinite. in a time frame and it means every node must spend The assumptions, which have been mentioned above, are considerable amount of power in every time frame. Another valid for most of the sensor networks. There are some important question about PACT is time synchronization. Every additional assumptions specifically for MIL-MON. TDMA based system must synchronize time and PACT does Sink node has high range transmitter as well as low range not propose a system for time synchronization and it does not transmitter. In this way, sink can use its low range transmitter take into account time synchronization overhead. to communicate with its neighbor nodes and it can send broadcasts for all nodes. All the sensor nodes can receive broadcasts of the sink. III. MIL-MON SYSTEM MECHANISMS Sink node has GPS. In this way, clock drift of sink is near zero . A. Overview C. Basic Mechanisms MIL-MON is proposed to monitor a relatively large area against intruders and send the data about intruders to sink as There are some basic mechanisms for operating MIL-MON soon as possible. The main design considerations of properly. These are time synchronization, distributed time MIL-MON are to be able to operate in large areas, to minimize scheduling and topology construction mechanisms. power consumption, to reduce delay. 1) Time Synchronization In order to satisfy design considerations, MIL-MON is According to assumption of this sensor network, every node designed as a TDMA based multihop sensor network. The sink in the system can receive the signals of the sink. Sink transmits of MIL-MON is supposed to have a high range transmitter so a broadcast signal to sensor nodes at the beginning of each that every node can receive its broadcast signal. time frame. These broadcast signals synchronize the network. MIL-MON includes global time synchronization, distributed This time synchronization scheme is very similar to LEACH, time slot assignment, topology construction mechanisms. In however there are a few differences based on the differences of addition to these basic mechanisms, there are data slot system assumptions of MIL-MON and LEACH. The number indication mechanism to save more power and rescheduling of cluster in LEACH is a piori in the system. Each cluster head mechanism to reduce delay. Before introducing the synchronizes its own cluster. MIL-MON has only one cluster mechanisms of MIL-MON, a sample application and basic and the head of cluster is the sink. Only the sink synchronizes assumptions of MIL-MON is presented. the whole sensor nodes with its broadcast signal. 2) Distributed Time Scheduling Mechanism (DTSM) B. Sample Application and Basic Assumptions Almost all sensor network architectures that use TDMA The potential applications of wireless sensor networks are produce its time schedule centrally. Cluster head collects the highly varied. Environmental monitoring, condition based data about its sensor nodes and produces the time schedule of maintenance, smart spaces, military, precision agriculture, its cluster. Time schedule is sent to nodes by the cluster head. transportation, factory instrumentation, inventory tracking are The main assumption of this system is that when cluster head just some of the sample application areas. Behaviors of one transmits a signal, sensor nodes can receive or vice versa. specific sensor node can be totally different from the others However, in MIL-MON, although the signal of the sink can even under the same conditions. In this case, it is clear that be received by all sensor nodes, the sink can not receive the 4 signals of all sensor nodes. In this case, sensor nodes can not its predecessor. The simplest choice is the closest predecessor send their join request signal. If distributed time scheduling candidate. In the next time frame, it sends its own algorithm is used, there is no need to send a signal from sensor advertisement signal. If it receives hop number as h from its node to the sink directly. predecessor, it advertises its own hop number as h+1. In the In centralized time scheduling algorithm, cluster head same signal, it also sends the node number of its predecessor. collects all the requests from sensor nodes in a certain This signal is received by the predecessor. If the predecessor protocol. In most of the systems, this protocol is contention receives a hop advertisement signal with its node number, it based. If the number of sinks is limited and number of sensor understands that the owner of the signal has become its child. nodes for each sink is very high, the traffic from sensor nodes In the third time frame, it listens to its all neighbors to learn to sink will be heavily loaded. Contention based protocols can which nodes became its children. Because in the third time not achieve high efficiency under heavy traffic. This can be a frame, the children of the sensor nodes advertise their hop serious problem for power sensitive systems. However, in number and predecessor node number. distributed time scheduling algorithms, there is no need for Figure 1 shows a sample run of the distributed topology communication between nodes and the sink directly. This construction algorithm. Suppose that A has caught an leads to power saving. advertisement from X with hop number h. It advertises its own According to DTSM, after sensor nodes are deployed, every advertisement signal with the parameters X and h+1. B catches node selects a random time slot as its own slot. It listens to all this signal and other advertisement signals, if any. If A is the time slots in the first time frame and only in its own slot, it closest node among the nodes that have sent advertisement transmits a special signal. If it receives a jammed signal, it signal, B sends its advertisement signal with parameters A and means there is a collision at that particular slot. The node h+2. collects all the collision slots. In the next time frame, it transmits a signal at the collision slots. In the same time frame, A receives adv(Y,h) from X it listens to its own slot. If it receives a signal, it means it has A sends adv(X,h+1) the same slot with another node’s slot so that their signals are B receives adv(X,h+1) jammed. In this case, it sleeps. If it does not receive any signal A receives adv(A,h+2) and understands that B is its child. B sends adv(A,h+2) and at its own slot, it can use that slot and it continues to operate. advertises that A is its This protocol is simple and consumes low power. However, predecessor. it does not result in a complete solution. Some nodes have to sleep. Fortunately, most of the time, sensor nodes are deployed Fig. 1. A sample run of topology construction algorithm. densely and the non-existence of a small number of nodes can be tolerated. If maximum hop number of the network is H, constructing 3) Topology Construction the topology takes H+1 time frame for the whole network. It is After distributed time scheduling, the second step for the fast, simple and scalable. It consumes low power and it results proposed sensor network is topology construction. Many in minimum hop paths, which is a good approximation for TDMA based sensor network systems use star topology. All minimum power consumption. the nodes are directly connected to cluster head. However, in 4) Rescheduling large areas, multi hop structure consumes less power than one Although power consumption is generally the most hop. There is a trade off between multihop and one hop important design issue in sensor networks, there are some structures. If the distance is less than 30 meters, direct other design considerations like delay. In some applications, transmission is more efficient than multihop transmission. delay between time of event and time of data arrival to the sink However if it is longer than 30 meters multihop strategy can is not crucial. However, if sensor network is designed for save energy . military purposes, delay can be very important. In a military If sensor network is aimed to cover a large area, average monitoring system, the existence of intruder should be distance between sensor nodes and the sink is supposed to be reported as soon as possible. much longer than 30m. In this case, multi hop strategy is an If sensor node is h hop away from sink, and one time frame essential part of MIL-MON architecture. is t seconds, the worst delay is t*h and t*h delay can not be A distributed topology construction algorithm is proposed to acceptable in some cases. Suppose that t is one second and support multi hop strategy. Time scheduling algorithm is also sensor node that has sensed the arrival of intruder is 7 hops distributed. In this way, the only scalability constraint of the away. If the average distance between neighbor nodes is 20 m., system is the transmission range of the sink. intruder is 140-150 m. away to the sink and the sink can know The basic structure of the algorithm lies on handshaking the existence of intruder after 7 seconds which can be enough signals. After getting a proper time slot, sensor nodes listen to for the intruder to harm the soldier that holds the sink. their neighbors to catch hop number advertisement signal. At Reducing delay is possible by the help of assigning time the beginning, only the sink sends hop number advertisement slots carefully. The rule is that smaller hop numbered nodes signal. Hop number of the sink is zero. If a sensor node can should get higher slot numbers. For example, if a sensor node, catch advertisement signal or signals, it chooses one of them as 5 say A is 3 hops away and its slot number is 320, slot number Node A Neighbors of A of another node with 2 hop number, say B, should be greater than 320. In this case, node A can send its message in 320 th Gets an adv. that includes hop number slot and node B can receive at 320th slot. In the same time of sender. Calculates its sub time frame, Receives new time slot number and checks frame, node B can send the data of node A to its predecessor. selects a time slot randomly and whether there is sends its time slot number to its collision. In order to realize to reschedule, time frame is divided into neighbors. If there is collision, it m sub time frames. If the whole time frame has n slots, a sub If it gets a signal from sends a signal to A. neighbors, it sleeps. time frame has n/m slots. The slot number assigned to a node If it does not get, it If there is no collision, it remains silent. with hop number h, must be in (m-((h-1) mod m))th sub time changes its time slot and continues with topology frame. According to this formula, the nodes with 1 hop number construction. will have a slot number from mth sub time frame. The nodes Fig. 3. Rescheduling mechanism. with 2 hop number will have a slot number from m-1th sub time frame. In this way, the slot number of consecutive hop An example helps to understand rescheduling clearer. Let us numbered nodes will belong to consecutive sub time frames. assume that the nodes in Figure 4 are one hop away from its Figure 2 shows the structure of this approach. consecutives. In this particular network, time frame has 300 time slots. The first slot is reserved for time synchronization one time frame divided into broadcast of the sink. Let us assume that DTS_SN algorithm m sub time frames has been run and the assigned time slots are as in the Figure 4(a). Sub time frame numbers mth (m-1)th ……. 1st Hop numbers 1 2 ……. m Sink, (2) A, (82) B, (91) C, (198) D, (245) Fig. 2. Rescheduling structure. There are m sub time frames in one time frame, and each sub time frame has n/m slots, if total number of slots in time frame is n. (a) In order to assign proper time slots, node must know its hop distance to the sink. In this case, rescheduling algorithm can Sink, (2) A, (250) B, (142) C, (45) D, (245) not be run before distributed time scheduling algorithm. Rescheduling should be placed between DTSM and topology (b) construction. Fig. 4. Example network and time slots (a) Before rescheduling. (b) After After running distributed time scheduling algorithm, non rescheduling. sleeping nodes has valid time slot. Nodes listen to its neighbors to receive an advertisement signal which includes According to rescheduling algorithm, the sink transmits a the hop number and node number of the sender. When a node broadcast signal with hop number 0. A receives this and receives a hop advertisement signal, rescheduling is started. calculates its sub time frame and selects a random time slot in The advertisement signal includes the hop number of the that particular sub time frame. In this example, A must get a sender. In this way, the receiver of the advertisement signal time slot between 300 and 201. Let us assume that A selects can calculate its hop number. In addition to this, node can get 250th time slot. It advertises its new time slot in 82nd time slot the number of sub time frames m from the sink’s (its current time slot. The sink and node B listens to 82 nd time synchronization signal. The node can calculate the sub frame slot for new time slot of A. If the sink and the nodes with hop number with the formula (m-((h-1) mod m)). It selects a number 2 listen to all its neighbors and collect the newly random time slot in the (m-((h-1) mod m))th sub time frame and requested time slots. If they determine a collision for new time sends a broadcast signal that includes the information about slots, they send a warning message in the slots that request the the selected time slot. The node sends this information in its new time slots. In this example, neither the sink nor B receives original time slot that was assigned with DTSM mechanism. a collision for the new time slot of A. Node A does not All the neighbors of the node listen to its signal and check receives a warning signal at 82nd time slot. In this case, new whether there is collision about its new slot number or not. If time slot number of A becomes 250. After rescheduling there is a collision, the nodes that catch the problem send a algorithm, topology construction mechanism is run with the signal in their own time slots. The node that tries to reschedule new time slots. So, A sends an advertisement signal with hop listens to its all neighbors in the next time frame. If it gets a number 1. The same procedure is run for all nodes. An signal from the others, it sleeps. If it does not get any signal example resulting network is presented in Figure 4(b). related with its new time slot, it means there is no problem and The relay of an event from D to the sink takes 737 time slots it starts to use its new slot and continues with topology for non-rescheduled network in Figure 4(a). However, it takes construction procedure. Figure 3 shows the signal handshaking only 305 time slots for the rescheduled network in Figure 4(b). required for rescheduling. The figure starts with the node that has got a slot with DTSM mechanism. 6 IV. PERFORMANCE RESULTS number of nodes and slots. The increasing rate is higher for It is very difficult to analytically model the interactions of MIL-MON with 500 time slot. Sleeping node rate is sensor nodes, even for the limited number of nodes. In order to acceptable for 1000 and 1500 time slots. However, sleeping investigate performance results of MIL-MON, simulation node rate is high for 500 time slots. method is used. Although there are some network simulators The same cases are investigated with non-rescheduled 1000 that can simulate wireless networks, there is no built-in sensor time slots, resheduled with 5 sub-slots, 10 sub-slots and 20 network module. In this paper, a new simulator has developed sub-slots. The results are presented in Figure 6. to simulate MIL-MON. Figure 6 shows that rescheduling increases sleeping node Simulation parameters are listed in Table1. Energy ratio linearly. consumption model is the same as in . Energy of a node is assumed to be supplied with 15 mg. Ni-Cd battery which can 10 support 2 J . 9 % of sleeping nodes 8 TABLE I 7 1000 SIMULATION PARAMETERS 6 1000-5 5 Parameter Default values 4 1000-10 3 1000-20 Power needed for radio 50 nJ electronics per bit 2 Power for receiving per bit 50 nJ 1 Power for transmitting per bit 50nJ + 10pJ*d*d (d is distance) 0 Max. range of nodes 30 m. 2 2,5 3 3,5 4 4,5 5 Power in one node 2J Number of nodes (x1000) Simulation area diameter 1000 m. Position of the sink Center of the area. Time for one time slot 1 ms. Fig. 6. Sleeping node ratio for different number of nodes with rescheduling. Bit rate 1 Mbps. B. Network Lifetime Performance results are discussed in three domains. These are performance of DTSM, network lifetime and delay. Network lifetime is one of the most common performance metrics of sensor networks. However, network lifetime A. Performance of DTSM definitions vary. In this paper, network lifetime is the time that DTSM is a distributed time scheduling mechanism to assign the first sensor node exhausts its energy. Network lifetime of proper time slots to sensor nodes. However, DTSM can not MIL-MON is discussed with the system load. In this assign time slots for every nodes. The nodes that can not get experiment, number of nodes is 2000, number of slots is 1000. time slot sleep and do not join into network. The ratio of The results are in Figure 7. sleeping nodes is an indicator of performance of DTSM. MIL- In Figure 7, system load is the number of events in one MON is simulated for 1000 slotted time frames. 1000 slotted second. If there is no event, the first node exhausts after more MIL-MON is investigated for non-rescheduled, rescheduled than 35000 seconds. Network lifetime decreases with the with 5 sub-slotted, 10 sub-slotted and 20 sub-slotted versions. increasing of system load. MIL-MON is designed to be used Figure 5 shows percentage of sleeping nodes for different for military purposes and the attack of intruder is not so number of nodes and different number of slots. Percentage of common in the battlefield. MIL-MON can operate for more sleeping nodes increases linearly with the increasing of than 10 hours. MIL-MON can be used for securing the environment especially for one night operations. 12 40000 % of sleeping nodes 10 35000 8 500 30000 25000 Time (s) 6 1000 20000 4 1500 15000 2 10000 0 5000 2 2,5 3 3,5 4 4,5 5 0 Num ber of nodes (x1000) 0 0,25 0,50 1 System Load Fig. 5. Sleeping node ratio for different number of nodes. Fig. 7. Network lifetime for different system loads. 7 C. Delay Rescheduling improves delay very successfully. Delay in Delay is very important for military applications. Long rescheduled with 20 sub-slot is 18 times smaller than non- lifetime is not enough, if it gives information too late. MIL- rescheduled MIL-MON, especially in the 10th region. (the MON uses TDMA and constructs a tree. It means delay of border of the sensor network area.) events is a function of distance from event to sink. Delay is investigated for different distance regions. Network area is V. CONCLUSION divided into 10 regions. The first region is 10m. distant from In recent years, sensor network is one of the hottest topics in the sink. The second is between 10-20m. distant from the sink, wireless communication area. In this paper, a new TDMA and so on. Figure 8 shows average delay of events for different based sensor networks for military monitoring (MIL-MON) is time slot numbers in these regions. proposed. The most important design objectives of MIL-MON are to prolong network lifetime, to reduce delay and to be able to operate in large areas. 20000 In order to realize MIL-MON, distributed time scheduling, 18000 Normal 500 topology construction and rescheduling mechanisms are 16000 Normal 1000 studied. Simulation results have shown that network lifetime 14000 Normal 1500 Delay (ms) 12000 of MIL-MON is long enough to be used as military 10000 monitoring. Delay problem sourced form TDMA structure 8000 can be handled by rescheduling mechanism and delay is 6000 reduced by up to 18 times. 4000 When a MIL-MON node exhausts, the network system is 2000 0 broken in this version of MIL-MON. However, sensor nodes 1 2 3 4 5 6 7 8 9 10 are prone to failure and sensor network should be able to fix Region number itself, when a node fails. As future work, a maintenance algorithm will be developed. Fig. 8. Delay for different regions for different number of slots. REFERENCES When distance between the sink and event increases, delay  Deborah Estrin, Lewis Girod, Greg Pottie, Mani Srivastava, increases dramatically. MIL-MON is an intruder detection “Instrumenting the World with Wireless Sensor Networks”, International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2001), Salt system and early alarm is critical. MIL-MON should give Lake City, Utah, May 2001. information as soon as possible, even if event is far from the sink. Even 500 slot is not acceptable. If intruder is 500m.  I. F. Akyildiz and W. Su and Y. Sankarasubramaniam and E. Cayirci. “Wireless Sensor Networks: A Survey. Computer Networks” (Amsterdam, distant and if MIL-MON gives information after 5 seconds, it Netherlands: 1999), 38(4):393-422, 2002. may be too late. Rescheduling is proposed to reduce delay. The same  E. Shih, S.-H. Cho, N. Ickes, R. Min, A. Sinha, A. Wang, and A. experiment is repeated with 1000 slot non-rescheduled, 1000 Chandrakasan, “Physical layer driven algorithm and protocol design for energy-efficient wireless sensor networks”, In Proc. ACM MOBICOM, pages slot rescheduled with 5, 10 and 20 sub-slot MIL-MON. The 272--286, Rome, Italy, July 2001. results are in Figure 9.  W. Heinzelman, A. Chandrakasan, and H. Balakrishnan, “An Application-Specific Protocol Architecture for Wireless Microsensor 14000 Networks”, IEEE Transactions on Wireless Communications, Vol. 1, No. 4, Normal 1000 October 2002, pp. 660-670. 12000 Re 1000-5  K. Sohrabi, J. Gao, V. Ailawadhi, and G. J. Pottie, “Protocols for self- 10000 organization of a wireless sensor network”, IEEE Personal Communications, Re 1000-10 Delay (ms) vol. 7, pp. 16 - 27, October 2000. 8000 Re 1000-20 6000  . Kottapalli, A. S. Kiremidjian, J. P. Lynch, E. C., T. W. Kenny, K. H. Law, Y. Lei, “Two-tiered wireless sensor network architecture for structural 4000 health”, SPIE'03. 2000  Guangyu Pei and Charles Chien, “Low power TDMA in large wireless 0 sensor networks”, MILCOM 2001 - IEEE Military Communications Conference, no. 1, October 2001, pp. 347 - 351 1 2 3 4 5 6 7 8 9 10 Num ber of region  W. Heinzelman, “Application-Specific Protocol Architectures for Wireless Networks”, PhD Thesis, Massachusetts Institute of Technology, June 2000. Fig. 9. Delay for different regions for different number of slots.  Wendi Heinzelman, Anantha Chandrakasan, and Hari Balakrishnan, “Energy-Efficient Communication Protocols for Wireless Microsensor Networks”, Proc. Hawaaian Int'l Conf. on Systems Science, January 2000. 8  J. Mannermaa, K. Kalliomaki, T. Mansten, and S. Turunen, “Timing performance of various GPS receivers” In Proceedings of the 1999 Joint Meeting of the European Frequency and Time Forum and the IEEE International Frequency Control Symposium, pages 287–290, April 1999.  Rex Min and Anantha Chandrakasan, “Energy-Efficient Communication for Ad-Hoc Wireless Sensor Networks”, 35th Asilomar Conference on Signals, Systems, and Computers, vol. 1, November 2001, pp. 139-143.  J. Frieman. “Portable Computer Power Sources” , In Proceedings of the Ninth Annual Battery Conference on Application and Advances, pp. 152-158, January 1994.