A hybrid multi-hop mobility assisted heterogeneous energy efficient cluster routing protocol for Wireless Ad hoc Sensor Networks

Abstract

Wireless Ad hoc Sensor Network (WASN) is becoming a demandable and evolving area of engineering which has been currently engaged in various fields particularly to make wireless communication system effective and reliable. WASN is resolving real time applications in the places where topographical, ecological terrains and atmospheric phenomenon are ever fluctuating. In this paper, a Hybrid Multi Hop Mobility assisted cluster routing protocol has been proposed for maximizing the system performance and throughput under heterogeneous real time mobile sensor network environment, by uniformly distributing energy among the sensor nodes. The behavior of node mobility here is being used to manage power consumption in cluster routing process. This can help researchers to project and design efficient energy aware wireless routing applications like tracking and data capturing in those areas where the network topology is ever changing. The proposed protocol can makes the communication network capable of routing packets to the base station in adverse and dynamic conditions with prolonged network operation.

Keywords

Wireless Ad hoc Sensor Network heterogeneity mobility energy efficiency clustering routing

Abbreviations

WASN

Wireless Ad hoc Sensor Networks

Cluster Head

Base Station

MSN

Mobile Sensor Node

MCH

Mobile Cluster Head

SSN

Static Sensor Node

SCH

Static Cluster Head

Gateway Node

LND

Last Node Died

FND

First Node Died

1. Introduction

Wireless communication is becoming a fundamental important factor devoid of which, envisioning the world is quite difficult. Wireless Ad hoc Sensor Network takes the collective and comprehensive features of Wireless Sensor Network (WSN) with ad hoc networks. Ad hoc sensor networks are planned with autonomous sensors communicating through radio without any prebuilt infrastructure. Recent advancement in wireless communications and electronics has empowered the engineers to develop low-cost, low power, multifunctional miniature devices for use in real time and remote sensing kind of applications. Sensor networks are composed of a large number of sensor nodes which consist of sensing, data processing and communication capabilities. In ad hoc networking scenarios it is tough to forecast the instance to deploy the nodes and which random topology the network will follow in the next instance due to impulsive node movement. Sensor networks are mostly data centric rather than address centric.

Incorporating realistic behavior in the current WASN helps to draw a distinct line between dynamic and fully dynamic states in terms frequent node displacement under proactive, reactive as well as hybrid network scenarios. In typical network setting each node requires a power source to do all device operations including data capturing, processing, trans-receiving etc. These applications of wireless sensor network essentially require a battery to be mounted on each sensor node. However, battery replacement is a challenge. Hence, battery consumption required to be smart enough to make every single as well as the whole network to be durable in terms of energy. Therefore, it is crucial to monitor and manage power exhaustion uniformly across the network, to get the best use of network.

Any technology that is in course of its development, offers a lot of challenges. In the same way, WASN also brings a lot of challenges and issues towards its improvement. Sensing, processing and packet communication by minute sensors with power restriction is not that simple task. To maximize WASN life extent, it is required to give emphasis on such algorithms, protocols and circuitries that can bring maximum functionality out of inadequate power source. Particularly, in wireless multi hop network scenario; for effective performance; its protocol has to act very competently in parallel with any change in network settings. Several protocols have been developed to deal with power problem in sensor networks. Most of the routing algorithms that are available for wireless and ad hoc networks can be categorized into three types i.e., direct transmission algorithms, hop to hop transmission algorithms and cluster based algorithms in both proactive and reactive network paradigm. Data aggregation and data fusion method have been used in clustering to inspire non redundant packet transmission at quicker pace to the Base Station with lower routing overhead.

As WASN, has been aimed for locations with zero human interference where geographical changes are continuous and dynamic. Hence, it is always desirable for an effectual routing protocol that has low transmission overhead and well organized data aggregation techniques to prolong the network operation by saving limited energy of the sensors uniformly. However, several cluster based energy efficient routing algorithms in WSN are available. To bring more energy efficiency in real time applications of WASN, existing cluster routing process requires modification for better capabilities under heterogeneous mobile version of the network.

The rest of the paper as follows, Section 2 brings knowledge about the previous work done in the related field. In Section 3, proposed approach has been described including Algorithm. Section 4 has presented the result and discussion. Conclusion of the paper is written in Section 5.

2. Related work

Clustering for energy preservation is proven effective for WSN [10,11,27,41]. Conceptually, clustering along with election of cluster heads to carry out data aggregation which indirectly improves the bandwidth consumption and life time of network [3,23]. Considering cluster based algorithms, several protocols have been developed where each of them having different attributes and enrichments. LEACH [26,34], a proactive cluster based probabilistic routing protocol has given adequate ground to think about novel energy aware strategies in wireless networks to overcome different network issues and challenges by providing energy efficiency to the network by means of clustering and routing which helped in designing protocols in new challenging network environments. In that light, authors in [16,17], have applied an energy based approach for CH selection before routing of packets by calculating the remaining energy as prime parameter. In [19], the proposed MODLEACH as an enhancement of LEACH protocol by minimizing network energy consumption by efficient CH replacement. In [7], author has proposed an energy efficient scheme that considers the cost of nodes for transmitting the data packets to the sink, where node cost considers both the remaining energy of the node as well as energy efficiency for multi hop communication in WSN.

In [18], TL-LEACH (Two Level hierarchy LEACH) was proposed to restrict direct communication to the BS, the protocol builds different levels of packet communication. This performs better energy management by distributing the energy load among the nodes, which increase the working span and quality of the network as compared to conventional LEACH. In [23], authors have given the concept of inter cluster communication to transmit fused data from CH to BS via. Multiple hopes i.e. from one CH to another and finally to BS, it would further enhance network life time. In [6], author has discussed a distributed multi-hop hierarchical clustering technique i.e. based on proximity information to quantify energy consumption in WSN. In [42], a hierarchical, distributed, energy-efficient clustering protocol HEED has been presented that differentiate intra cluster and inter cluster communication and regulate the packet transmission with the use of dual power levels at each sensor nodes. In [1], a hybrid approach has been proposed by combining the characteristics of flat multi-hop routing and hierarchical multi-hop routing to minimize the total power consumption in the network which results in decreased traffic and efficient data compression.

TEEN [20] and APTEEN [21] are proposed as reactive protocols after LEACH, that conserves extra energy wastage by introducing the concept of dual thresholds called hard threshold and soft threshold in data communication, that gives good result in network life time by showing reactive nature. These thresholds can be applied and used in different phases of routing protocols to enhance the performance with respect to network utility. Decentralized Reactive Clustering (DRC) is another reactive protocol has been proposed in [38] where the process of clustering started only in the occurrence of events in the sensor coverage area. DRC uses a power control technique to reduce energy usage in cluster construction. Clustered Aggregation (CAG) [40] was proposed to operate on spatial correlations that exist between the sensory data to decrease the number of communications by producing approximate outcomes to aggregate queries.

As WASN is a kind of ad hoc network; its performance and quality of service evaluation cannot give realistic result under static network environment. In reality the ad hoc behavior of sensor nodes in application domain encourage node displacement dynamically at different time instance becomes more supportive to construct application specific networks. Extensions have been made in several existing protocols under mobility but as the area of study is vast hence, many challenges still persist to deal with such dynamic network construct. LEACH-Mobile [15] is an extension of LEACH for mobile centric environment where the nodes are mobile in steady state. LEACH-Mobile Enhanced [39] protocol was proposed which gives emphasis on successful data transfer even though nodes are moving. LEACH-ME considers and addresses the situations where an energy rich CH may come out from cluster due to mobility but it will not affect much in the performance due to the use of group mobility model which is less random by nature. M-LEACH uses attenuation model [28] is based on transmission distance and the Radio Frequency (RF) attenuation exponent which varies with the network environment quality. Several other protocols have been proposed under mobile network environment like, DECA (Distributed Efficient Clustering Approach) [43] is a distributed energy aware clustering protocol under multi hop mobility scenario, where each node at times transmit hello packets by calculating a score based on residual energy, connectivity and identity metrics. Distributed Efficient Multi hop Clustering (DEMC) [2] is an extension of DECA [43], where it uses recovery technique that reduces the packet loss during inter cluster communication with decreased number of transmission and increased packet delivery ratio under mobility sensing environment. In [29], author has presented the protocol where a packet is sent through multiple hops at cluster level by allowing each sensor to make individual decisions regarding its mode of operation to prolong the network’s lifetime by minimizing the average energy spent for each communications performed in the mobile network environment. Most of these protocols have been simulated and studied under limited node mobility by implementing group mobility model. These protocols address data communication under mobility of nodes in a particular pattern instead of random patterns and do not establish any link between the different movement patterns of nodes to quantify energy consumption in communication.

Stable Election Protocol (SEP) [33] is a heterogeneous clustering protocol that assigns weighted probability to each node becoming a Cluster Head (CH). In DEEC [30], residual energy in each node at a particular instance becomes the election criteria for a node to be a CH. In [14], a new protocol is proposed that works better in heterogeneous wireless networks in terms of network stability and life time without much affecting the performance metric. In [35], author states that different heterogeneous routing protocols for WSN and in most of the network environment heterogeneity comes based on energy levels of the nodes which helps in selecting CHs. In [13], a new heterogeneous wireless sensor networks (HWSNs) model has been proposed with both energy heterogeneity and computational heterogeneity where the CH election algorithm is based on the energy dissipation forecast method which considers the residual energy and energy consumption rate in all nodes. It has been understood that heterogeneity among nodes comes; when nodes distinguish themselves based on functionalities resulting differences in their energy level at a particular time instance, but has not been addressed heterogeneity in terms of node mobility in WASN. These heterogeneous protocols also have missed to discuss about the energy consumption and system throughput with respect to heterogeneity in the scale of mobility. Success of WASN can be counted if its performance will be decent in realistic network environment for which it has been actually designed.

Energy is also a vital parameter in secure network applications. In [22], the author has put forth the different mechanisms that are dealing with energy awareness and security in network construct. The possibility of security outbreak is always more in case of wireless network. There are several severing battery drain attacks in wireless networks where energy itself is an entity to protect and is considered as an indicator to reveal those security outbreaks. Therefore, security protocols need to be energy aware [22]. Most of the security protocols in wireless networking are associated with routing called secure routing in consideration of energy consumption. Recent advancements unveil that a network with Intrusion Detection System (IDS) can make substantial burden in the network traffic level with packets inspection at different points in the network. In [24] the author has focused on an energy aware approach to distribute packet inspection over the network nodes based on early or late discovery and discard of rogue packets to perform network energy savings. In [25], author has discussed the energy impact of an IDS enabled Internet Service Provider Network (ISPN) multi-hop network. In that approach the author has carried out the energy leakage assessment [25] in terms of early bad packets detection under both centralized and distributed IDS model. Energy leakage can be a vital parameter to model any secure energy efficient network.

To construct communication in mobile WASN by using traditional clustering approaches are little difficult due to rapid topology change. So it is required to study and develop new approaches to deal with such quick network changes. Hence, these protocols need to be tested and simulated under realistic as well as extreme dynamic condition. To simulate and model such realistic mobile ad hoc environment for sensor network, several mobility models and their characteristics [4,5,8,9,12,31,32] have been studied to develop heterogeneous movement patterns based on parameters like location, velocity, speed, accelerations etc., in the sensor nodes. The author in [36] has proposed and compared gateway selection methods based on signal strength among clusters in Wireless Ad hoc Networks. The primary goal is to design a gateway based energy aware multi-hop cluster routing protocol in a heterogeneous mobile WASN environment. In multi hop cluster based hierarchical mobile wireless environment, the routing of packets to the intended sink nodes is a tough task and the limited power backups in the sensors urge the design process of routing protocols to be energy efficient. The details of the proposed network model have been described in the following section.

3. Proposed network model

As mobile network environment is always dynamic and complex than static one, hence the designing of protocol need to be hybrid, simple and come up with all good features that brought up from different protocols and it should be acceptable in terms of cost, complexity and maintenance. To measure energy efficiency and system throughput, respect to heterogeneity in terms of mobility in WASN requires several improvements and modifications. To implement this, following detailed key conceptual insinuations have been considered in the proposed WASN protocol.

The networked nodes are maintaining random mobility pattern. Heterogeneity in terms of mobility has been added. That is, randomly some nodes become mobile, partly mobile and immobile in every round as shown in Fig. 1(a). As described in the Fig. 1(b) there are two clusters intersecting each other representing the movement of both MSN and MCH node randomly and simultaneously, Cluster 1 has a SSN and Cluster 2 has SCH. The numbering signifies the random movement of networked node within a predefined time bound at auto generated velocity with a pause at every new location. Because of node movement, at any instance a node may come in the intersection range of radio sphere and stop or pause there, then it behaves as GN between the nearby multiple reachable clusters. The random displacement and movement incorporated in the nodes can be summarized as per the following set of Eqs (1), (2) $\begin{array}{l} (1) & S = S_{min} + (S_{max} - S_{min}) * rand, \\ (2) & T_{2} = T_{1} + din / S + Pt, \end{array}$ where, S is the speed of random movement of sensor node; $din$ is the intermediate distance of two distinct waypoints; $T_{2}$ , $T_{1}$ are different time instances of intermediate waypoints; $Pt$ is the pause time at every distinct intermediate waypoint during complete random movement.

Fig. 1.

Heterogeneous mobile WASN (a) heterogeneous mobility; (b) mobility with clustering.

Due to frequent and random movement of nodes, each time there is a good chance of penetration of MCH of one cluster to come inside another cluster, it is termed as MCHnew in the new cluster. In such situation based on energy level, either MCHnew or CH wins the race for that instant in that cluster. If MCHnew wins, then old CH supress its CH status and announces the CHnew ID to all its members currently in its range as in Fig. 2. As nodes are heterogeneous in their displacement hence there is always a chance of unequal clustering at any network instances.

It has been observed that mobility brings significant changes in network construct and in heterogeneous mobile environment the nodes which makes higher number of random displacements, due to rebroadcasting its presence and urge of association and re-association with nearby cluster; at each new waypoint locations results significant consumption of energy as shown in Fig. 3. The lines coming out of nodes are the sense of broadcasting and rebroadcasting their presence in respective locations. This activity in this version of the WASN brings instability in terms of energy in subsequent epochs.

This phenomenon has been minutely described in the form of mobility count with new set of equations (as shown in Eqs (3), (4)) later has been used to derive the probabilistic threshold for the proposed model. $\begin{array}{l} (3) & M_{n} = \sum_{k = 1}^{m} 1, \\ (4) & M_{n} \propto E_{cn}, \end{array}$ where, $M_{n}$ signifies the total number of nth node movement; m signifies the mobility count for nth node which varied randomly; k is variable for m; $E_{cn}$ signifies the energy consumption of nth node at any point of time.

Fig. 2.

CH movement across the clusters.

Fig. 3.

Broadcasting and re-broadcasting of mobile nodes.

Under mobility, it has been observed that, nodes are selected to be CHs, come nearby to each other; out of them one will becoming CH for that region for current instance and other behave like normal nodes but keep their CH status. To conserve energy with respect to round those special status nodes can be reused in further rounds as they are with full energy or above appropriate energy value required to work as CH node. These nodes will not participate in new CH election process in the current round. This reduces extra energy consumption and uniform utilization of node energy in mobile WASN environment.

The new mobility based metric has been calculated and multiplied in each round for each node in probabilistic threshold calculation as described in Eqs (5), (6), (7), which makes clustering process mobility dependent. Equation (5) calculates the mean of total node movements whereas Eq. (6) calculates the maximum time node displacement followed by Eq. (7) as new calculated threshold. $\begin{array}{l} (5) & M_{mean} = \frac{\sum_{1}^{n} M_{n}}{n}, \\ (6) & M_{r} = max [M_{n} | \forall n \in N, \forall r \in r_{max}], \\ (7) & T (n) = (\frac{P}{1 - P * (r mod \frac{1}{P})}) * (\frac{M_{mean}}{M_{r}}) . \end{array}$

Due to random and heterogeneity in mobility, the multiplied mobility metric never 1 and more the node movement in the network less chance of becoming CH for the next round due to excess energy depletion in pause and broadcast.

To make uniform distribution of energy usage both energy threshold and mobility metrics are taken in to consideration for efficient clustering.

Each node performs intra and inter cluster communication at different power level based on difference in packet travel distance, so that efficient communication can be possible by expending appropriate energy which will be helpful in checking energy usages.

To conserve energy the concept of dual threshold is used in intra cluster communication only not in inter cluster communication, so that when the sensed data satisfies both the thresholds, then only the sensor will activate all its processes to make intra cluster communication, which results less flow of redundant data throughout the network and sensor can enjoy more time to be inactive and conserve more energy.

Multi hop communication via intermediate CH or Gateway Nodes has been considered to check direct and long range communication to BS that, expends less energy.

Normally, in conventional WASN nodes closer to BS overhear due to over traffic load in that region results early power depletion as shown in Fig. 4. To minimize the communication overhead in this energy prone region, if any sensor node at any instance become closer to the BS where communication link is good then it has to transmit directly without using the energy of any intermediate nodes.

Fig. 4.

Concentering circular coronas around the sink [37].

Finally to measure the system performance in term of packets, the system throughput is calculated as per the following equation (8) for this heterogeneous mobile WASN as shown in Fig. 11. $\begin{matrix} (8) & {ST}_{put} = P_{Rcvd} / (R_{C} * N_{alive}), \end{matrix}$ where, ${ST}_{put}$ is the system throughput; $P_{Rcvd}$ is the packet received out of packet transmitted to BS; $R_{c}$ is the current round number; $N_{alive}$ is the number of alive node in the round $R_{c}$ .

The proposed algorithm has been discussed in details as follows:

Algorithm.

Initialization of Sensor Parameters for nodes (N) in accordance with 1st order radio-energy model; Set, Min_Speed, Max_Speed and Time of node Mobility;

Random deployment of $nodes (i)$

Calculate the $dead_nodes (i)$ and $\sum Energy (i)$

$if (time (i) < = Time)$

Calculate, $distance_m (i)$ , $velocity (i)$ and $Speed (i)$ ;

Perform, Heterogeneous random mobility on sensor $nodes (i)$ with random $pause (i)$ ;

Broadcast: $hello_JOIN (i)$ ;

Calculate: $\sum pause (i)$ ;

Calculate: $Energy (i) : = E_total - (Energy (i) * \sum pause (i))$ ;

[Calculate Energy remains and dissipated]

end if

$if (node (i) = = CH & & Energy (i) > E_threshold)$

Select such $nodes (i)$ as CH for current round and perform the data communication to BS.

$Count (CH)$ ;

Calculate: dist_threshold

$CH (i)$ calculate self-distance $distance (i)$ from BS.

$CH (i)$ perform control_packet transmission based on, DIST_threshold_1;

$CH (i)$ make Packet transmission expend $Energy (i)$ ;

[Calculate Energy remains and dissipated]

else

GOTO: Step 6

end if

$if (Count (CH) \neq Required (CH) & & Energy (i) > 0)$

[Probabilistically, perform CH election process for current round]

Calculate: DIST_threshold_1

Generate: $node (i) \to random (0, 1)$

Calculate: new Mobility assisted threshold $T (n)$ , $\begin{matrix} T (n) = (p / (1 - p * \mod (r, round (1 / p)))) * (M_{mean} / M_{r}) \end{matrix}$

$CH (i)$ calculate self-distance $distance (i)$ from BS

$CH (i)$ perform control_packet transmission based on, DIST_threshold_1;

$CH (i)$ make Packet transmission expend $Energy (i)$ ;

[Calculate Energy remains and dissipated]

end if

[Intra Cluster Reactive Communication]

At, $time (i)$ : $nodes (i)$ associate by $hello_JOIN (i)$ to different $CH (j)$ based on MIN (CH_distance_INtra);

Control packet communication to $CH$ ,

TDMA schedule allocated for $time (i)$

$if (sense_new (i) > threshold_HARD&&sense_DIF > threshold_SOFT)$

$node (i)$ intra-cluster communication and transfer packets to associated CH at $time (i)$ ;

$node (i)$ regulate power for short range intra-communication DIST_threshold_2,

$node (i)$ , consume energy for packet transmission;

$CH (i)$ : $DATA_aggregarion (i)$

[Calculate Energy remains and dissipated]

end if

[Inter Cluster Muti-Hop Proactive Communication]

TDMA schedule allocated for $time (i)$

$if (node (i) = = CH (i))$

At, $time (i) : CH (i)$ get associate with different [ $Count (CH) - 1$ ] based on MIN (CH_distance_INter);

Calculate: distance to receive and aggregate x-bit of data from $CH (i)$ to $HEAD_CH (i)$

$if (CH_distance_INter > DIST_threshold_1) CH (i)$ , consume energy for packet transmission;

else

GOTO: Step 9

end if

[Calculate Energy remains and dissipated]

[CH-Gateway-CH Proactive Communication]

$if (CH_distance_INter > DIST_threshold_2 &&CH_distance_INter < DIST_threshold_1)$

$if (S_Strength (node (i) to CH (i)) > S_StrengthTH&&Energy (i) > 0)$

$GWY_eligible : = {node (i)}$ ;

$GWY (i) : = Select (GWY_eligible)$ ;

end if

do, CH-CH and CH-Gateway-CH communication;

Calculate: control_packet(Drop);

packet(Rcvd);

node(alive);

SYS_Troughput(N);

exit

The proposed algorithm begins with the initialization of sensor parameters with required sensor mobility attributes in Step 1. In Step-2 of the algorithm the nodes are randomly deployed to form preliminary ad hoc sensor network setting. Step-3 calculates the dead node and energy level of whole network at any instance of time for each round. In Step-4 new heterogeneous mobility has been incorporated at each individual node level by checking their movement pattern, velocity, speed with respect to time instances and random pause pause(i) among the nodes to maintain heterogeneity in real time time(i). This step calculates the energy wastages of individual nodes based on frequency of broadcasting of hello_JOIN(i) message. Steps 5 and 6, responsible for conditional dynamic CH election process by using E_threshold as ( $0.5 * E_{initial}$ ) followed by mobility metric assisted probabilistic threshold $T (n)$ based CH election process. This process minimizes the energy consumption overhead by uniform utilization of CH energy with required number of CH at every round which ultimately checks over election of CH like in conventional cluster routing system. Step 7, forms intra-cluster reactive communication by associating nodes to CH with respect to time instances time(i) and perform packet transmission by mapping the threshold_HARD and threshold_SOFT value with new sensed value and once it satisfy the dual threshold test then the new data of a particular current region will be transmitted to CH at time stamp time(i). Step 8 forms inter-cluster proactive multi-hop communication by associating nearby CHs and there is no conditional dual thresholds for inter CH transmission. Step 9, segregate the long range multi-hop transmissions into short range transmissions by electing and selecting suitable gateway node GWY(i) for time(i) for packet routing to reach nearby CH. Each sensor nodes regulates their power consumptions by switching their radio transmitter in between two power level based on two distance threshold DIST_threshold_1 and DIST_threshold_2 for inter (i.e. CH-to-CH or CH-Gateway-CH) and intra cluster communications respectively that gives energy efficiency to the protocol. Where, DIST_threshold_1 is calculated as $sqrt (Efs / Emp)$ and DIST_threshold_2 is calculated as $sqrt (0.1 * (Efs / Emp))$ . Step 10 calculates the system throughput $SYS_Troughput (N)$ by considering the successful packet transmitted and received at BS with respect to number of alive node and round.

4. Simulation results and discussions

The following simulation parameters have been considered as in Table 1 to model the ad hoc network of sensors in MATLAB2013a for carrying out the simulation. In this section, the network performance has been described based on following evaluation parameters.

Network lifetime: It is the time interval from the start of the network operation till the last node die.

Cluster count: The number of Cluster formation per round till the network is alive.

Packets to BS: The total number packets generated and to be transmitted to the BS per round.

Packets through CH: Packets transmitted through CH per round.

Network Energy Level: The Network’s residual energy per round in order to analyse the energy consumption pattern of the whole network. Higher remaining energy ensures graceful degradation of network life.

Success Rate: Out of total packets generated towards the BS via multi-hop communication, how many packets able to reach their intended destinations in the routing path towards the BS per round.

System Throughput: System throughput signifies the numbers of data packets received successfully at the BS by the remaining alive nodes in the network in each round.

Table 1
Simulation parameters

Simulation parameter Value

Network size in terms of area 400 × 400 sq. m.

No. of mobile nodes 100 nos

Probability (p) 0.1

Initial energy ( $E_{initial}$ ) 0.5 J

Energy dissipated (Trans receiving) 50 nJ/bit

Amplification energy (Inter Cluster and CH to BS) 10 pJ/bit/m²

Amplification energy¹ (Efs) $10 * 0.000000000001$

Amplification energy² (Emp) $0.0013 * 0.000000000001$

Energy (data aggregation) 5 nJ/bit

Control packet size 1 bit

Data packets size 4000 bit

Min. speed of nodes 0 ms

Max. speed of nodes 20 ms

Pause time 2 ms

Time 50 sec

Simulation parameter	Value
Network size in terms of area	400 × 400 sq. m.
No. of mobile nodes	100 nos
Probability (p)	0.1
Initial energy ( $E_{initial}$ )	0.5 J
Energy dissipated (Trans receiving)	50 nJ/bit
Amplification energy (Inter Cluster and CH to BS)	10 pJ/bit/m²
Amplification energy¹ (Efs)	$10 * 0.000000000001$
Amplification energy² (Emp)	$0.0013 * 0.000000000001$
Energy (data aggregation)	5 nJ/bit
Control packet size	1 bit
Data packets size	4000 bit
Min. speed of nodes	0 ms
Max. speed of nodes	20 ms
Pause time	2 ms
Time	50 sec

LEACH, is one of the popular cluster routing protocol in WSN that has been framed for static environment. But, in reality the network condition in WASN may not be static. Hence, under the first phase of this test, LEACH has been tested under real time heterogeneous mobility scenario by implementing random way point mobility model and termed as $LEACH (M)$ . The objective behind using this model is to test the behaviour of the network in the scenarios where the topology change is dynamic and unpredictable. To enhance the performance of $LEACH (M)$ , the above algorithm along with new network implications as per the proposed network model has been simulated. The proposed protocol i.e. Hybrid Multi Hop Mobility Assisted Cluster Routing Protocol outperforms $LEACH (M)$ in terms of different network characteristics including system throughput.

As discussed in the previous sections, the proposed protocol has come up with different modified features of several proactive as well as reactive protocol features that makes it hybrid. One of the ideal motivating factor behind the proposed work is to use the power of mobility to manage power consumption in the form of calculated threshold.

Fig. 5.

Number of dead nodes with respect to rounds (a) static environment; (b) real time hybrid heterogeneous multi hop mobile environment.

In Fig. 5(a) the protocol LEACH has been tested in terms of dead nodes with respect to round under static environment. But, when random heterogeneous mobility has been applied on this, its performance degrades drastically as shown in Fig. 5(b) for $LEACH (M)$ . However this has been enhanced in the proposed protocol in terms of dead node count with respect to number of rounds. It has been observed due to realistic and dynamic mobility pattern, a node that moves beyond the control of the radio range of any other nearby node then that node becomes the victim of early expiry, and this possibility can hardly be avoided in WASN environment with such kind of random node mobility pattern, as described in Fig. 5(b). Hence, to measure the lifetime of these kinds of network, it is most suitable to scale the efficiency of the networks by considering LND instead of FND. Though, very few nodes have lost their energy heavily in the beginning but with the passage of time and rounds the network behavior gradually changes towards increased efficiency as compared to $LEACH (M)$ . Use of gateway nodes along with multi-hop communication minimizes the long range communication into multiple short range communications to reduce the energy depletion at node level.

Figure 6(a) shows LEACH protocol in terms of number of clusters with respect to round in static environment, but when tested in heterogeneous mobile environment its performance lowers considerably in $LEACH (M)$ as shown in Fig. 6(b). However, in the same network environment the behavior of the network in terms of clustering with respect to number of rounds has been boosted up by applying the proposed protocol as described in the same figure. The objective of this study unveils the fact that in a hierarchical network both in static as well as dynamic environment, the degree of data communication and routing depends on network’s CH count. By introduction of real time random mobility in the conventional network, the cluster head count drastically decreased which ultimately hampers the network communication and connectivity, resulting in poor performance. In that state the proposed protocol enhance the cluster head count and constantly maintain this parameter until the network is alive and last packet reaches the BS.

Fig. 6.

Number of cluster formation with respect to rounds (a) static environment; (b) real time hybrid heterogeneous multi hop mobile environment.

Fig. 7.

Number of packet transmitted to BS with respect to rounds (a) static environment; (b) real time Hybrid multi hop heterogeneous multi hop mobile environment.

In Fig. 7(a), the static protocol has been tested in terms of number of packets transmitted to BS with respect to round, but when tested in heterogeneous mobile environment as $LEACH (M)$ , the performance has been dropped as shown in Fig. 7(b). However, with efficient clustering as discussed in dynamic network environment in the proposed work, the limitations of packet drop and packet retransmission has been minimized to large extent resulting in increased number of generated packets being transmitted to BS with respect to number of rounds as compared to $LEACH (M)$ , as shown in the same Fig. 7(b).

In Fig. 8(a), the static protocol has been tested in terms of number of packets transmitted through CH with respect to round before reaching to BS, but when tested under the simulated dynamic heterogeneous environment its performance decreases noticeably in $LEACH (M)$ as in Fig. 8(b). As $LEACH (M)$ does not support multi hop communication at CH level hence it has been taken in terms of number of packets transmitted to BS through CH with respect to number of rounds. However, in proposed one the graph has been shown for the packets transmitted through CH-to-CH multi hop communication which makes the graph to remain lower as compared to $LEACH (M)$ as shown in the Fig. 8(b). The objective of this study is to show that the proposed approach filters and checks the redundancy in packet transmission up to BS via. multi-hop communications between CHs.

Fig. 8.

Number of packet transmitted through CH (a) static environment; (b) real time hybrid multi hop heterogeneous multi hop mobile environment.

Figure 9 displays the average residual energy of network per round. This signifies the continuous energy depletion of the whole network with simulation time respect to rounds, In this study each node owns 0.5 joule energy in the beginning. Hence, total 100 sensors will bring 50 joule of energy for the entire network.It has been clearly visualised that the proposed protocol performs better in terms of energy conservation in individual nodes as well as the total network as a whole.The network efficiency and effectiveness will be more if the network life span is more. Because, network life span depends on the energy level of alive nodes available in WASN at any instance of time. Hence, as discussed earlier, LND criteria is more appropriate for lifetime assessment of such kind of networks under mobility.

Fig. 9.

Network energy level with respect to rounds.

As it is acceptable that in between wired and wireless medium the strength of successful packet routing is weaker than that of conventional wired communication due to poor link strength, loss of acknowledgement, packet drop etc. However along with heterogeneous mobility the scenario becomes critical due to unpredictable packet loss and piculiar node displacement with time. Here, it has been shown that Fig. 10 signifies the success rate of both $LEACH (M)$ and the proposed protocol in terms of total packets generated and transmitted towards the BS with respect to the actual packets routed and finally expected to be received at BS. This has been presented as Rcvd packet in Fig. 10 under multihop network scenario. In addition to the loss in data transmission in mobile environment, higher success rate of the network can be viewed as better utilization of energy and communication links. Since nodes are moving around dynamically,the chance of failure in packet transmission is more. However in the proposed protocol, it has been observed that under the heterogeneous random mobility of nodes the average success rate is nearly 66% with extended network life.

Fig. 10.

Success Rate in terms of packet routed with respect to rounds.

Fig. 11.

System throughput.

Figure 11 describes the system throughput. As per Eq. (8), system throughput is calculated based on packets received with respect to the total alive nodes in current round. As discussed earlier Fig. 5(b), the number of dead nodes increases slowly as compared to $LEACH (M)$ with respect to the rounds resulting better alive node count in the network. Under the simulated network environment System throughput has been calculated in terms of average data packets that are able to reach to the BS by the proposed cluster routing protocol under heterogeneous mobile WASN compared to $LEACH (M)$ . The mobility in real wireless environment lowers the whole system throughput due to significant packet loss and drop during frequent transmissions as well as retransmissions. It has been observed that, the cause of early energy drain of the node in random locations in the network is due to frequent join and detachment of nodes inside the clusters which results in more control packets retransmission as compared to data packets in between clusters. To trace out new nearby CH, most of the time nodes keep their radio transceiver ON. These issues have been taken care in proposed protocol to enhance the system throughput compared to $LEACH (M)$ . More alive nodes retain significant network energy that encourages higher throughput of the network and minimum energy consumption contributes enhanced network lifetime.

5. Conclusion

In this paper, Hybrid Multi hop Mobility Assisted Cluster Routing protocol has been proposed for performance enhancement in terms of energy efficiency and system throughput. Its performance has been assessed under heterogeneous real time mobile environment. The heterogeneous movement of nodes changes the activity of the network that leads to early discharge of the battery as compared to conventional one. To survive the network in such real time network settings, the proposed protocol minimizes the number of dead nodes with reasonably balanced clustering with respect to the number of rounds by using the new network mobility metrics based probabilistic threshold. This leads to extra power restoration and more non redundant packets routing with enhanced system throughput as compared to conventional network system. This protocol can maintain uniform network utilization with extended network operation in critical real time application domains and minimizes network congestion.

Footnotes

Acknowledgements

We thank Science and Engineering Research Board (SERB), Department of Science and Technology (DST, Govt. of India) for providing research support.

References

A.E.A.A.

Abdulla,

Nishiyama and

Kato, Extending the lifetime of wireless sensor networks: A hybrid routing algorithm, Computer Communications 35(9) (2012), 1056–1063.

Ali and

Madani, Distributed efficient multi hop clustering protocol for mobile sensor networks, The International Arab Journal of Information Technology 8(3) (2011), 302–308.

L.M.C.

Arboleda and

Nasser, Comparison of clustering algorithms and protocols for wireless sensor networks, in: Proceeding of Canadian Conference on Electrical and Computer Engineering (CCECE 2006), Ottawa, ON, 2006, pp. 1787–1792. doi:10.1109/CCECE.2006.277358.

Aschenbruck,

Gerhards-Padilla and

Martini, A survey on mobility models for performance analysis in tactical mobile networks, J. Telecommun. Inform. Technology 2 (2008), 54–61.

Bai and

Helmy, A survey of mobility models in wireless ad hoc networks, in: Wireless Ad hoc and Sensor Networks, Kluwer Academic Publishers, 2004, pp. 1–30.

Banerjee and

Khuller, A clustering scheme for hierarchical control in multi-hop wireless networks, in: Proceedings of 20th Joint Conference of the IEEE Computer and Communications Societies (INFOCOMS’01), 22–26 April 2001, Anchorage, AK, pp. 1028–1037. doi:10.1109/INFCOM.2001.916296.

Bhattacharjee and

Bandyopadhyay, Lifetime maximizing dynamic energy efficient routing protocol for multi hop wireless networks, Simulation Modelling Practice and Theory 32 (2013), 15–29. doi:10.1016/j.simpat.2012.11.006.

Bouaziz and

Rachedi, A survey on mobility management protocols in Wireless Sensor Networks based on 6LoWPAN technology, Computer Communications, in press, Corrected proof, available online 22 October 2014. doi:10.1016/j.comcom.2014.10.004.

Camp,

Boleng and

Davies, A survey of mobility models for ad hoc network research, Wireless Communications & Mobile Computing, Wiley, Special Issue on Mobile Ad hoc Networking: Research, Trends and Applications 2(5) (2002), 483–502. doi:10.1002/wcm.72.

10.

C.-Y.

Chong and

S.P.

Kumar, Sensor networks: Evolution, opportunities and challenges, Proceedings of the IEEE 91(8) (2003), 1247–1256. doi:10.1109/JPROC.2003.814918.

11.

Das,

Barman and

J.D.

Sinha, Energy efficient routing in wireless sensor network, Procedia Technology 6 (2012), 731–738. doi:10.1016/j.protcy.2012.10.088.

12.

Doci,

Samakovitis and

Raju, Impact of mobility in ad hoc protocol design, in: Proceeding of IEEE WRI World Congress on Computer Science and Information Engineering (CSIE’09), 31 March, 2009–2 April, 2009, Los Angeles, CA, IEEE, 2009, pp. 38–43.

13.

Haibo,

Yuanming and

Guangzhong, EDFM: A stable election protocol based on energy dissipation forecast method for clustered heterogeneous wireless sensor networks, in: Proceedings of the 5th IEEE International Conference on Wireless Communications, Networking and Mobile Computing, 2009, pp. 3544–3547. doi:10.1109/WICOM.2009.5304152.

14.

Halder and

Das Bit, Enhancement of wireless sensor network lifetime by deploying heterogeneous nodes, Journal of Network and Computer Applications 38 (2014), 106–124. doi:10.1016/j.jnca.2013.03.008.

15.

D.-S.

Kim and

Y.-J.

Chung, Self-organization routing protocol supporting mobile nodes for wireless sensor network, in: Proceedings of the IEEE First International Multi-Symposiums on Computer and Computational Sciences (IMSCCS ’06), Hanzhou, Zhejiang, 20–24 June 2006, IEEE, 2006, pp. 622–626. doi:10.1109/IMSCCS.2006.265.

16.

Kumar,

Prateek,

N.J.

Ahuja and

Bhushan, DE-LEACH: Distance and energy aware LEACH, International Journal of Computer Applications 88(9) (2014).

17.

C.-K.

Liang,

Y.-J.

Huang and

J.-D.

Lin, An energy efficient routing scheme in wireless sensor networks, in: Proc. of IEEE Advanced Information Networking and Applications-Workshops (AINAW’08), IEEE, 2008, pp. 916–921. doi:10.1109/WAINA.2008.199.

18.

Loscri,

Morabito and

Marano, A two levels hierarchy for low-energy adaptive clustering hierarchy (TL-LEACH), in: Proceedings of the IEEE 62nd Vehicular Technology Conference (VTC’05), 25–28 September 2005, IEEE, 2005, pp. 1809–1813. doi:10.1109/VETECF.2005.1558418.

19.

Mahmood,

Javaid,

Mahmood,

Qureshi,

A.M.

Memon and

Zaman, MODLEACH: A variant of LEACH for WSNs, in: Proceeding of Eighth IEEE International Conference on Broadband and Wireless Computing, Communication and Applications (BWCCA), 2013, pp. 158–163. doi:10.1109/BWCCA.2013.34.

20.

Manjeshwar and

D.P.

Agrawal, TEEN: A routing protocol for enhanced efficiency in wireless sensor networks, in: Proc. of 1st Intl. Workshop on Parallel and Distributed Computing, 23–27 April 2000, San Francisco, CA, USA, 2000, pp. 2009–2015. doi:10.1109/IPDPS.2001.925197.

21.

Manjeshwar and

D.P.

Agrawal, APTEEN: A hybrid protocol for efficient routing and comprehensive information retrieval in wireless sensor networks, in: Proceedings of the International Parallel and Distributed Processing Symposium (IPDPS’02), Ft. Lauderdale, FL, USA, 15–19 April 2002, IEEE Computer Society, Washington, DC, USA, 2002, p. 48. doi:10.1109/IPDPS.2002.1016600.

22.

Merlo,

Migliardi and

Caviglione, A survey on energy-aware security mechanisms, Pervasive and Mobile Computing, Special Issue on Secure Ubiquitous Computing 24 (2015), 77–90. doi:10.1016/j.pmcj.2015.05.005.

23.

Mhatre and

Rosenberg, Design guidelines for wireless sensor networks: Communication, clustering and aggregation, Ad hoc Networks 2(1) (2004), 45–63. doi:10.1016/S1570-8705(03)00047-7.

24.

Migliardi and

Merlo, Improving energy efficiency in distributed intrusion detection systems, Journal of High Speed Networks 19(3) (2013), 251–264. http://dl.acm.org/citation.cfm?id=2595811.

25.

Migliardi and

Merlo, Modeling the energy consumption of distributed IDS: A step towards green security, in: Proceedings of the 34th IEEE International Convention MIPRO (MIPRO’11), Opatija, pp. 1452–1457. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=5967289&tag=1.

26.

Nayebi and

Sarbazi-Azad, Performance modeling of the LEACH protocol for mobile wireless sensor networks, Journal of Parallel and Distributed Computing 71(6) (2011), 812–821. doi:10.1016/j.jpdc.2011.02.004.

27.

Nazir and

Hasbullah, Energy efficient and QoS aware routing protocol for clustered wireless sensor network, Computers & Electrical Engineering 39(8) (2013), 2425–2441. doi:10.1016/j.compeleceng.2013.06.011.

28.

L.T.

Nguyen,

Defago,

Beuran and

Shinoda, An energy efficient routing scheme for mobile wireless sensor networks, in: Proceeding of the IEEE International Symposium on Wireless Communication Systems (ISWCS’08), 21–24 October 2008, Reykjavik, pp. 568–572. doi:10.1109/ISWCS.2008.4726120.

29.

Papadopoulos,

Navarra,

J.A.

McCann and

C.M.

Pinotti, VIBE: An energy efficient routing protocol for dense and mobile sensor networks, Journal of Network and Computer Applications 35(4) (2012), 1177–1190. doi:10.1016/j.jnca.2011.05.004.

30.

Qing,

Zhu and

Wang, Design of a distributed energy-efficient clustering algorithm for heterogeneous wireless sensor networks, Computer Communications 29(12) (2006), 2230–2237. doi:10.1016/j.comcom.2006.02.017.

31.

Santi, Topology control in wireless ad hoc and sensor networks, ACM Computing Surveys 37(2) (2005), 164–194. doi:10.1145/1089733.1089736.

32.

Silva,

Sa Silva and

Boavida, Mobility in wireless sensor networks – Survey and proposal, Computer Communications 52(1) (2014), 1–20. doi:10.1016/j.comcom.2014.05.008.

33.

Smaragdakis,

Matta and

Bestavros, SEP: A stable election protocol for clustered heterogeneous wireless sensor networks, in: Proc of the Int’l Workshop on in Second International Workshop on Sensor and Actor Network Protocols and Applications (SANPA’04), 2004, pp. 251–261.

34.

Tanwar,

Kumar and

J.J.P.C.

Rodrigues, A systematic review on clustering and routing techniques based upon LEACH protocol for wireless sensor networks, Journal of Network and Computer Applications 36(2) (2013), 623–645. doi:10.1016/j.jnca.2012.12.001.

35.

Tanwar,

Kumar and

J.J.P.C.

Rodrigues, A systematic review on heterogeneous routing protocols for wireless sensor network, Journal of Network and Computer Applications 53 (2015), 39–56. doi:10.1016/j.jnca.2015.03.004.

36.

Wu,

Kim and

Park, A gateway selections using signal strength among clusters in ad hoc networks, in: Proc. of 7th International Conference on Information Security and Assurance, Vol. 21, ASTL, 2013, pp. 180–185, onlinepresent.org/proceedings/vol21_2013/46.pdf.

37.

Xiaobing,

Guihai and

Sajal, Avoiding energy holes in wireless sensor networks with non-uniform node distribution, IEEE Transaction on Parallel and Distributed Systems 19(5) (2008), 710–720. doi:10.1109/TPDS.2007.70770.

38.

Xu and

Qi, Decentralized reactive clustering for collaborative processing in sensor networks, in: Proceedings of the 10th International Conference on Parallel and Distributed Systems (ICPADS’04), 7–9 July 2004, IEEE, 2004, pp. 54–61. doi:10.1109/ICPADS.2004.1316080.

39.

G.S.

Kumar,

M.V.

Vinu Paul and

K.P.

Jacob, Mobility metric based LEACH-Mobile protocol, in: 16th IEEE International Advanced Computing and Communications Conference, ADCOM’08, December 14–17, 2008, IEEE, 2008, pp. 248–253. doi:10.1109/ADCOM.2008.4760456.

40.

Yoon and

Shahabi, Exploiting spatial correlation towards an energy efficient clustered aggregation technique (CAG), in: Proceedings of IEEE International Conference on Communications (ICC’05), 16–20 May 2005, IEEE, 2005, pp. 82–98. doi:10.1109/ICC.2005.1495034.

41.

Younis,

Munshi,

Gupta and

S.M.

Elsharkawy, On efficient clustering of wireless sensor networks, in: Proceeding of Second IEEE Workshop on Dependability and Security in Sensor Networks and Systems (DSSNS 2006), 24–28 April 2006, Columbia, MD, IEEE, 2006, pp. 78–91. doi:10.1109/DSSNS.2006.8.

42.

Younis and

Fahmy, HEED: A hybrid, energy-efficient, distributed clustering approach for ad hoc sensor networks, IEEE Transactions on Mobile Computing 3(4) (2004), 660–669. doi:10.1109/TMC.2004.41.

43.

Yu,

J.H.

Li and

Levy, Mobility resistant clustering in multi-hop wireless networks, Journal of Networks 1(1) (2006), 12–19. doi:10.4304/jnw.1.1.12-19.