Routing in a Wireless Sensor Network using a Hybrid Algorithm to Improve the Lifetime of the Nodes

A reliable and secure routing protocol for Wireless Sensor Networks should be easy to maintain, reliable and cost-efficient. In this paper, we propose a hybrid routing algorithm using Ant Colony Optimization and Minimum Hop Count scheme. The proposed hybrid methodology provides an optimal routing path that ensures balanced and minimal energy consumption. The hybrid algorithm is unique in maintaining network topology, balancing network load and searching for the optimal route. The proposed algorithm with the WSN model is implemented in C++ and the simulation output proves our algorithm to outperform other similar routing algorithms.The proposed work indicates animprovement in network lifetime, success rate in finding the best solution and rate of convergence. The existing techniques are reviewed and their strengths and weaknesses are diagnosed and compared with our proposed hybrid methodology that integrates the strengths of both the algorithms.


Introduction
One of the basic requirements of the Wireless Sensor Networks (WSN) is their lifetime. A number of methods have been proposed over the years for increasing the WSNs' lifetime, paying attention to details such as device control, topology management, routing, data processing and device placement. Of these criteria, routing plays a crucial role in increasing the lifetime using an energy efficient mechanism. However, there are a number of challenges to be addressed which emerge due to large scale of network, dynamic nature of the network [8], inherent unreliability in communication and energy constraint on the nodes. Hence ease of maintenance, high reliabilty and low cost are the key aspects that are required to ensure optimal routing protocol in WSN. When compared with other algorithms, WSN routing using Minimum Hop Count (MHC) as well as WSN routing with Ant Colony Optimization (ACO) [10] are said to be the optimal method used to identify the shortest route. Hence we have proposed a hybrid algorithm which integrates these two routing algorithms, taking into consideration dynamic network topology, network load balancing and energy constraints in the nodes. The following are the basic algorithms of operation used by the proposed hybrid algorithm: 2. Aneighbouring hop which has a lower count will be preferred when compared with the node itself, following MHC algorithms' hop count-classification.
3. Nodes that are farther away are chosen to transfer data, providing a means to choose the optimal path.
4. Success rate in searching for the optimal path is improved using the mutation strategy introduced. Taking battery depletion into consideration, use of multiple paths is encouraged.
5. The proposed work uses a dynamic energy threshold strategy wherein energy consumption of the individual nodes are taken into consideration to stop the depletion of certain nodes.
6. Load balancing is achieved by equalizing the workload of data transmission.
This paper provides a review of the two algorithms involved namely multiple hop count and ant colony optimization,in Section 2. This is followed by Section 3 which gives a description of the proposed model. Section 4 gives an extensive simulation results and a conclusion is drawn in section 5.

Related Work
Over the years, the behaviour of fishes, bees and ants have been used in swarm intelligence (SI) to successfully develop and apply ways to solve major optimization issues as per  [5], Jose et al. [6], Stutzle and Hoos [7] and Dorigo and Gambardella [8] research, it has been found that ACO has been successful in solving many industrial and scientific optimization process. During the past decade a number of challenges in WSN have been addressed by ACO based routing algorithm. Since energy is one of the crucial criteria that has an impact on the performance of WSN. Multipath Routing Protocol (MRP) based ACO was proposed in [9] which uses multiple paths to locate the optimal nodes which have minimum energy consumption. Another algorithm, Breadth First Search was combined with ACO by Khoshkangini et al in [10] to find the shortest and best path which will decrease energy consumption as well as the transaction time.
In [11] Han et al. İntroduced the novel methodology of MHC which works on the basis of flooding algorithm and Directed Diffusion algorithm. Based on the hop count between the sink node and the source node, the transmission path is decided using this algorithm. The 2 major disadvantages when using this algorithm are as follows: 1. Though the MHC is able to find the optimal path, there is no reasonable methodology to pick the next node except to transfer data through the parent node thereby increasing the energy consumed and also increasing the redundant information.
2. In order to update the route, flooding is initiated by the sink node at regular intervals. This will result in energy overhead and will also increase the cost of the network due to redundant data transmission.
In order to address the issue of transmission routing and energy consumption, Ho et al. in [12] proposed the use of ladder diffusion algorithm based on ACO. This was further modified by Du et al. in [13] using an energy-aware ladder diffusion methodology which uses Ho's proposal and further solves 'energy pole' and 'hot spot' problems based on the residual energy of the node [14], [15] and [16].

WSN Network Model
A WSN comprises of a sink node along with many distributed wireless sensor nodes in an organised manner. The data is transferred from the target area till it reaches the sink node where it is processed further [17]. There are 6 assumptions in the proposed algorithm: 1. Data collection is done periodically by the sink node and the data is then forwarded based on routing.
2. Based on the distance, transmission power can be adjusted accordingly.
3. Link errors resulting in overtime retransmission and packet error are not taken into account. 4. It is not possible to replenish the sensor nodes from energy, but are constrained.
5. All the nodes present in the network will have same amount of initial energy, communication capability and computing power, making them homogenous.
6. In the monitoring area, the nodes are deployed in a random fashion. This location remains constant and will not change at any point of time. The sink node will be provided adequate knowledge of the location of the entire network.

Model of Energy Consumption
In general, the energy consumed by a node for the purpose of data communication is comparatively high

Hop Count Classification
The purpose of the proposed algorit Networks. A planar network topo algorithm [19]. First, the location classification is used to count the ho the hop count.
x If the hop count is h±2, the x If the hop count is h-1, then Thus this type of methodology will time complexity and accelerating commence on failure of a sensor n count classification network topolog this energy consumed [18]. Power amplifier energy re the two major parts which contribute to data tr d to decrease the distance such that it is lower than th for data transmission can be reduced such that as th ases, the energy consumed during data transmissi thm is to increase the lifetime of the nodes in a typic ology based hop-count classification network is u n information of every individual node is examin op values. Based on this algorithm, there are two alt h-hop node will not be able to transmit data. n the h-hop will be chosen. be able to pick the next hop node at a quicker pace, the convergence speed. Maintenance of the netw node or on detecting wireless link breakage. Fig.1. gy.

Dynamic Energy Threshold Strate
In this paper, we have implemented energy consumption for the WSN.T defined during initialization for all the energy threshold is determined. When the network is running, there node is below that of threshold, a ne

Convergence Characteristics C
It is important to find the optima topology and limited energy powe routing path, with high success r LTAWSN and ACO is compared w hybrid routing protocol requires less

Network Lifetime
Since most of the energy in a W Data transmission is possible only if information to the sink node. The li power. The network lifetime of the and Classic ACO in Fig.3. It is id trategy a dynamic energy threshold strategy which will be u The lower bound energy threshold and energy thre the nodes. In general, when an ant is searching for Based on it, the next hop will be fixed to be a node e is a constant consumption of energy. When the en ew route discovery process will be initiated by the sin tics Comparison al path with least cost calculation because of the er. Hence the proposed algorithm must excel in fi rate and convergence speed. The convergence sp with that of the proposed methodology and it is found ser number of iteration as observed in Fig.2.

g.2. Convergence Speed Comparison
SN is used for data transmission, network lifetime p f the nodes that are to forward data have enough pow fetime of a network is the time when the node is sai e proposed algorithm is compared with EALD, AC dentified that the proposed methodology will be ab useful in balancing eshold are a new hop node θ, on a priority order. nergy present in the nk node.

Conclusion
A hybrid routing protocol is propos dynamic networking capability, s consumption. Our proposed work pr Ant Colony Optimization and Minim has resulted in an overall improvem optimal path of routing, wehave im results indicate that the proposed m when compared to other algorithms. ork Lifetime Comparison of Various Algorithms ed in this work, taking into consideration the impro speed of convergence, balancing network load rovides an optimal routing methodology by combini mum Hop Count, which leads to an optimal solution ment in the performance and efficiency of the net mplemented the use of a dynamic energy threshold st methodology increases network lifetime and is also .