ECBK : Enhanced Cluster Based Key Management Scheme for Achieving Quality of Service

Wireless sensor networks consist of many small nodes with distributing devices to monitor conditions at different locations. Usually wireless sensor nodes are sprinkled in a sensor field grouping limited areas. This paper highlights the Enhanced Cluster Based Key management (ECBK) protocol to achieve secure data delivery based on clustering mechanism. This protocol gives more importance to Cluster Coordinator node, which is used to coordinate the members and provide protective communication among the sensor nodes to enhance reliability. In Enhanced Cluster Based Key management two types of nodes are deployed. The high power nodes form clusters with surrounding nodes to enable the routing process without interference. This paper introduces ECBK protocol that balances the load among the clusters, achieves high throughput, end to end delay will be reduced, routing overhead also reduced and also it prolongs the network lifetime. Simulation results show that the presence of high transmission nodes reduces the delay, load balance, routing overhead, and enhances the throughput increased by 45% compared to other similar methods.


Introduction
Wireless sensor network (WSN) is a collection of micro sensors that are powered by low energy batteries and equipped with micro processors, small memory and radio transceivers.Security of these networks emerges as a critical issue [1].Without proper security, it is impossible to trust [2]- [4] the results collected from sensor networks [5].Main security threats in WSN are node Sybil attack, compromise attack, code injection attack Hello flood attack and selective forwarding attack.Security mechanisms and key management schemes are used to detect threats to individual sensor nodes [6]- [8].To design secure wireless sensor networks, security measures may be incorporated in sensor nodes.
Sensor nodes trusts on battery power providing.Their communication capability and energy storage capability are very restricted [9].Hence efficient utilization of node energy, balancing network energy consumption and extension of network lifetime has become a primary design objective for wireless sensor network.Clustering is proved to be an efficient scheme for growing the lifetime of WSNs [10] [11].In a clustering scheme the nodes are divided into groups.They are mostly based on geographical properties.Each group has a single leader (Cluster Head) and several ordinary nodes (Member Nodes).A cluster head usually serves as a local coordinator for its cluster, playing intra-cluster transmission arrangement, data aggregation, forwarding [12] and so on.In the network model, gateway nodes connect between different clusters if there is no direct communication between the cluster heads [13].The objectives of the clustering are to decrease the total of the transmission power which is summed up over selected path of the nodes and to balance the load among the selected path of the nodes for growing the network lifetime [14] [15].
Enhanced Cluster Based Key management scheme for wireless sensor network can be used to increase the network lifetime in a secured way.These protocols have the Cluster Head Algorithm which is used to generate the coordinator node to increase the enhancement of the network lifetime of the Wireless Sensor Network.

Related Work
Clustered routing is suitable for networks which deploy large number of nodes and are highly scalable.In 2004, the first clustering algorithm LEACH was proposed for dropping power utilization.In LEACH, the clustering task is rounded among the selected paths of the nodes, based on time period.Cluster Head (CH) is used to communicate directly to forward the data to the Base Station (BS).Even as most efforts so far have focused on an energy-efficient clustering scheme [16], attention to the performance of multi-hop network was quite limited.In [17] hierarchical soft clusters with fuzzy enabling non-exclusive overlap clusters are implemented using a total version of time-division multiple access (TDMA).Although the clustering shows good performance, the sink is assumed to be reachable which may not be the case always.A new protocol LEACH-R planned in [18] improves the choice of cluster-head by considering the residual energy of the nodes throughout selection of cluster-head, thereby reducing the likelihood of low-energy nodes being selected.However, analysis of the cluster head trustworthiness has not been thought of.CLENER protocol [19] provides energy-efficiency by cluster formation using fuzzy logic and Cluster Head election based on a probability function.The drawback is that the base station has not been subjected to energy restrictions and is located inside the sensing field.Trust evaluation of Cluster Head has been successfully carried out in [20].Every Cluster Head reports its trust execution result to all other Cluster Heads to the BS that is infallible with all other physical security.A novel Cluster Based Routing Protocol (CBRP) for prolonged sensor network lifetime has been proposed in [21].In the set-up phase, clusters are generated followed by the steady-state phase where a routing tree is constructed and aggregated data are sent to the sink node.In CBRP protocol, each node maintains a table that contains the neighborhood information about its neighbors consuming time to build the routing tree and maintaining the table.In [22] a Clusterbased Energy-efficient Scheme (CES) for elect a cluster-head to evenly issue energy utilization in the overall network and therefore obtain a longer network lifetime has been proposed.Each sensor is responsible for maintaining a table called Table Cluster, in which information from the local cluster members is stored.Hence proper updating of table information becomes necessary.Cluster based routing protocol for heterogeneous WSN proposed in [23] minimizes the energy consumption and increases the network survivability.However, the nodes are assumed to be uniformly distributed in the network.All nodes can send data to the base station.Hence reliability and trust are not considered [24].Proposes Uniform Distribution Technique (UDT) for selecting CHs and to make a network throughout that every sensor remains within the transmission vary of CHs and so, the amount of the network is prolonged.This technique does not consider heterogeneous nodes.An energy-efficient scheme is usually evaluated by the network lifetime which is measured by the time that the first or the last node dies.Thus, network lifetime is tightly coupled with the network performance [25]- [28].
A Distribution Group Key theme for Wireless Sensor Networks has been proposed in the net of things situation.As an imperative a part of the web of Things (IoT), Wireless Sensor Networks (WSNs) need to be completely integrated into the web.Key cluster distribution scheme for WSNs in the IoT scenario organizes sensor nodes into groups.Each group has a hierarchical structure.The secure finish-to-end communication [29] protocol is employed to distribute cluster keys for subgroups to the trustworthy head nodes and also the head nodes then share out the cluster keys from end to finish underlying tree-based topology and wireless multicast to reduce energy utilization.Deals with a hybrid key management Scheme based on Wireless clustered Sensor Networks.Based on node's own location the weakness of session key is constructed.The use of hierarchical thinking reduces the amount of key storage and computing.The results of simulation show our proposed algorithm outperforms DSH (Dynamic Self-Healing) protocol [30] DD and LEACH in terms of conserving energy, security and prolonging WSN lifetime.

Network Model
To change the network model, we tend to affordable a few reasonable assumptions as follows.Figure 1 demonstrates the Network Model for the Proposed System. 1) There are n sensor nodes that are distributed randomly in an x, y square field.
2) All the nodes and the BS are stationary after deployment.
3) All the sensor nodes can be homogeneous except cluster head.4) Each node has an identity (id).5) Assume that the cluster coordinator node has the following properties: a) the base station sends query request tasks and broadcasts to the member nodes; b) it maintains the energy of each node, location and other information; c) it makes dynamic perception adding or removing nodes; d) it establishes virtual path among the cluster nodes and other cluster coordinator nodes for information transmission and maintains established routes;

ECBK Working Process
Trustworthy ECBK is used to improve the security of the network, by introducing a simple key management mechanism.Data transmission take place between the clusters and cluster members is completed through associate increased multi-path mechanism which additional improves the responsibility of data transmission and implements network load equalization.The cluster objectives area unit to minimize the overall transmission power aggregated over the nodes within the designated path and to balance the load among the nodes for prolonging the network life.Phase-I Cluster-head selection Sensor nodes deployed within the environment, all the nodes send their energy levels to the base Station.On the energy level and geographic area, a node with high energy will be selected as Cluster Head.

Crypto System
Security mechanisms and key management scheme detect threats to individual sensor nodes.To implement a secure connection between the nodes it requires the establishment of a key.Different types of keys are needed for different types of nodes.Cluster Head and Base Station stores preloaded master key whose main function is to prevent the network attack during initialization which causes information to be stolen and   generates the key pair which will be automatically sent to coordinator node.This phase is divided into three parts as follows.

The Establishment of a Key among Cluster Coordinator Nodes and Base Stations
Each coordinator node and base station generates a session key according as follows: ( ) where R is a random number for a base, with the coordinator node's private key.Base station key K m ( 1, 2, , m n =  ) encrypted obtaining ki R and sent to each coordinator node.Communication between coordinator nodes and base station in order encrypt to guarantee authentication should be deposited in the key.

Authentication of Nodes in Cluster
Each cluster member node for data transmission contains the same initial energy and has a unique ID number.A Cluster coordinator node is identified and maintained in each cluster.It monitors cluster node and helps in the nodes authentication to the Base Station.CCN is the cluster coordinator node and CCN  is the  is number of cluster ( 1, 2, , i n =  ) node, V represent maximum number of round, MACR represent as cryptographically certified MAC.A → B{C} indicates that A sends the message C to B, and ACK is the acknowledgement [15] [16].

The Key Establishment among the Cluster Heads
According to R ki and hash function  generate a unique key and key pair, such as coordinator node CCN ID for CCN identifier, and  ccn as its position information.Based on the two key, K ccn is generated as follows: ( )

Key Distribution and Establishment
Cluster coordinator nodes communicate with the Base Station to inform the number and location of the Cluster Head.Each Cluster Head CH i ( 1, 2, , i m =  ) stores allocation key share.Key segmentation process is summarized as follows.The base station first generates the function ENC and E will be encrypted as secret S, E = ENC k (s).Using decomposition algorithm divide E into m obtaining 1 2 , , , m  E E  E   , and assign to each cluster head.The key K is then divided into m obtaining 1 2 , , , m k k k  . Finally put two tuples ( ) as Key Shadow S i and send to the Cluster Head CH i [16] [18].Before communication, each kind of nodes also needs node-to-node authentication.If this succeeds, data transmission can be carried out.

Routing Protocol for Sensor Network
The following steps involved in the General routing-Best Forwarder selection 1) Find if Destination node is found among the neighbours.2) If found, set the next hop as Destination node and Exit.Else continue.
3) For each member -node among the neighbours, do the following: //Check for positive progress towards destination 4) Check if its distance from the Destination node is greater than or equal to the distance between the current forwarder and Destination node.If yes, break.Else, add node to an array.5) Find the nodes in the array, based on their distance to destination.6) If the node has reception power greater than the threshold value, choose that node as the Best Forwarder.
Else consider the next node in array.7) Exit.
Algorithm for cryptosystem 1) Generate session key ( ) by base station and co-ordinator node.
2) CCN  sends status request message as CCN  → BS {, REQ, R ki } to BS. , , , m k k k  and put the tuple as ( ) as key shadow S i then send to Ch i If these steps are succeeded then carried out the transmission.

Performance Evaluation
The proposed scheme is based on ECBK protocol.The main goal is to enhance the security of data transmission and prolong the network lifetime.Network parameters and values are given below (Table 1).
To evaluate the network performance, metrics such as load balancing, average end-to-end delay, network throughput and routing control overhead are considered.The proposed protocol is compared with Distributed Self-Healing protocol where no cluster formation is done, DD and LEACH.

Performance Metrics
Load balancing factor uses comprehensive weight value composed of distance between the head and the member and the residual energy to improve cluster member choice.
It also uses optimization threshold value to avoid load imbalance.B f to measure the balance degree of clusters is given by ( ) where  low is the number of nodes with energy lower than the average of the total number of nodes in the network,  is the number of cluster coordinator nodes,  is the total number of nodes,  −  is the number of participating nodes in the data transmission.The graph in Figure 2 shows the load balance factor of the proposed protocol.The simulation time is plotted along the X axis and load balance value is plotted along the Y axis.This graph shows that ECBK outperforms the other protocols in terms of load balance factor.ECBK chooses the node with high energy to become the cluster head.Since the Cluster head and routing path are chosen for each transfer, the energy levels of the nodes remain balanced.
Average end-to-end delay is the average of the sum of delays for transmission of packets from source cluster head to the destination given by where td is the data transmission delay of cluster heads.In Figure 3 the average number of time delays is plotted against end-to-end delay.The graph shows that the proposed protocol performs better than DCL and DD.However its performance is similar to LEACH during the start of the simulation but performs better than LEACH as time increases.In ECBK, nodes that make positive progress towards the destination are chosen.Hence path deviation is eliminated.High energy nodes tend to broadcast in a better manner.The above stated two reasons can be accounted for the less end to end delay.
Network throughput is the number of packets of successful transmission given by ( ) send s Tp n m P T = − where, time is denoted as   and  send is each cluster head that sends the packet number.The graph shown in Figure 4 is used to identify the network throughput in different time periods.The proposed protocol shows better network throughput compared to the other protocols.The network throughput of ECBK is highest compared to the other protocols, as seen in Figure 5 because high energy nodes are chosen as Cluster Heads and only nodes with the threshold reception power are chosen for intermediate hops.
Routing Control Overhead: overhead will increase the extent of disseminated unwanted data and redundant process at intermediate nodes further as base station where D C is the pay expenses which are cluster coordinator node controlling according to the built path transmission cluster heads need and T C is multiple paths in need of the overhead of data transmission.
Figure 5 shows the routing overhead in network.The proposed protocol performs far better than DSL and DD.As time increases it performs better than LEACH.Opportunistic routing done by ECBK is responsible for the reduced routing overhead.

Longevity
The numbers of alive nodes are calculated for several simulation periods.The graph in Figure 6 indicates that ECBK ensures more number of nodes alive for the different simulation times

Network Lifetime
As seen in Figure 7, ECBK prolongs the lifetime of the network due to increased longevity of nodes.High energy node selection for Cluster Head and opportunistic routing can be accounted for the better performance.

Conclusion
In this paper clustering with key management scheme ECBK is proposed for routing in WSN.ECBK helps to solve the problems encountered in data transmission security by introducing a simple key management mechanism, the secret sharing technology.In order to increase the reliability of the working node we also introduced the node evaluation mechanism is incorporated and low trust value nodes have been dropped.General routing-Best cluster selection algorithm is employed, which further improves the reliability of data transmission and implements network load balancing.The clustering objectives are to minimize the total transmission power of the selected path of the nodes and the balancing in the selected path of the nodes for increasing the lifetime of the network.Simulation tests demonstrate that ECBK performs much better than LEACH, DD and DHL in terms of load balance, average end to end delay, throughput and routing overhead.The simulation results proved that the ECBK is performed well compared to all the other Cluster based Scheme methods.Congestion control algorithms, mobile sink and proper scheduling methods can be incorporated into ECBK as future work.
The base station is well educated concerning the geographical locations of the nodes.Base Station can choose the Cluster Heads and multicast this data to them.Select Cluster Head (NetInfo) Node Ni send N i (EL i , L i ) to BS//El i is energy level of node, L i is Location of node If El i > e select i th node as CH//e is threshold value, CH is cluster head Else Ni is MN//MN is member nodePhase-II-Cluster formationAfter Cluster Head selection the cluster formation section starts.Each Cluster Head broadcasts its status as Cluster Head to the surrounding nodes.Nodes with Received Signal Strength Indicator (RSSI) worth response to this beacon message and show their temperament to affix this cluster.The node sends a request to the various Cluster Head and in response be a part of Confirmed Alert is shipped by the Cluster Head.In this approach clusters are formed with one Cluster Head each.One in every of the Cluster Member will act as a Coordinator node.It will not sense any data.It only coordinates and controls the Cluster Head's work such as selecting the direction of data transfers.
3) BS receives REQ message and find CCN  corresponding R ki replay as BS → CCN  {  , MAC  (ACK)} to CCN i 4) CCN  receives this message, calculate (ID i ) and send a message as CCN  → BS { * , ID i MAC(ACK)} to BS. 5) If BS receives this message then CCN  authentication is successful.