Security-Reliability Analysis of Multi-hop LEACH Protocol with Fountain Codes and Cooperative Jamming

In this paper, we investigate trade-off between security and reliability of Fountain codes (FCs) based low-energy adaptive clustering hierarchy (LEACH) networks, where the encoded packets are sent to the destination by using a cluster-based multi-hop transmission scheme with the assistance of cluster heads (CHs). With presence of an eavesdropper, a cooperative harvest-to-jam technique is employed to reduce the quality of the eavesdropping channels. Particularly, each cluster randomly selects a cluster node that generates artificial noises on the eavesdropper. For performance evaluation, we derive exact closed-form expressions of outage probability (OP) and intercept probability (IP) over Rayleigh fading channels. We then perform Monte Carlo simulations to verify the theoretical results, and compare the performance of the proposed scheme with that of the conventional LEACH scheme without using the jamming technique.


Introduction
Due to low battery power, limited memory, processing and communication capabilities, performance of wireless sensor networks (WSNs), i.e., lifetime, outage probability, error rates and channel capacity, is limited.For the energy-efficiency issue, low-energy adaptive clustering hierarchy (LEACH) protocol [1]- [3] can be used to prolong the network lifetime.In LEACH, the sensor nodes are grouped into local clusters in which cluster heads (CHs) are selected to manage their own clusters in a centralized manner, i.e., gathering the data from the cluster nodes and sending to the sinks or to the neighbour CHs.Moreover, dual-hop/multi-hop relaying protocols [4]- [7] are often used to enhance the data rate, extend the network coverage, mitigate effect of fading environments and compensate the performance loss due to low transmit power and hardware imperfection.
Security is also a critical issue in WSNs due to the limited computation, memory and storage.Recently, physical-layer security (PLS) [8]- [9] has been proposed to provide security for wireless communication systems.Due to the simple implementation, PLS is suitable for WSNs [6]- [7].To enhance secrecy performance, in terms of average secrecy capacity, secrecy outage probability and probability of non-zero secrecy capacity, the authors in [10]- [12] proposed diversity relaying protocols to enhance secrecy capacity via increasing the quality of the data links.In [13]- [14], joint cooperative relaying and jamming methods were proposed to further improve the secrecy performance.The basic idea of the cooperative jamming protocol is that friendly jammers cooperate with the legitimate receivers so that the generated interference only reduces the channel capacity of the eavesdropping links [15]- [16].Different with [10]- [16], references [17]- [19] investigated the trade-off between security and reliability by evaluating intercept probability (IP) and outage probability (OP) of the eavesdropping and data links, respectively.
Fountain codes (FCs) [20]- [21] have gained much attention due to the simple implementation.The transmitters employing FCs can transmit a limitless number of the encoded packets to intended receivers that can recover the original data with a sufficient number of the encoded packets.Due to broadcast of the wireless medium, the eavesdropper can easily receive the encoded packets and recover the data.Recently, FCs based secure communication protocols [22]- [25] were proposed and analysed.The authors of [22] considered the FCs based secure delivery protocol, where if the legitimate receiver can obtain sufficient number of the encoded packets before the eavesdropper, the data transmission is secure and successful.In [23], the FCs based dual-hop relaying protocol using the cooperative jamming technique was investigated.Reference [24] presented an efficient FCsbased multicast model to achieve security for Internet of Things (IoT) systems.In [25], performance of a downlink MISO scheme exploiting FCs, transmit antenna selection (TAS) and cooperative jamming was evaluated.
To the best of our knowledge, there has been no published literature related to the FCs based secure multihop LEACH protocol.In this paper, the source data that is encoded by FCs is sent to the destination via the intermediate clusters.To guarantee security for the data under attack of an eavesdropper, each cluster randomly selects a cluster node to realize the cooperative jamming operation.For performance evaluation and comparison, we derive exact closed-form expressions of IP and OP for the proposed protocol and the corresponding protocol without using the cooperative jamming technique, over Rayleigh fading channel.Finally, we perform Monte Carlo simulations to verify the theoretical analyses.
The rest of this paper is organized as follows.The system model of the proposed protocol is described in Section 2. In Section 3, the expressions of IP and OP are derived.The simulation results are shown in Section 4. Finally, this paper is concluded in Section 5.

System Model
Figure 1 presents the system model of the proposed protocol, where the source (CH of cluster 0) wants to transmit its data to the destination (CH of cluster M) via the multi-hop LEACH scenario with the assistance of multiple relays (CHs of the intermediate clusters).Using FCs, the source data is divided into L packets that are then encoded appropriately to generate the encoded packets [22]- [26].Next, the source transmits the encoded packets to the destination.Assume that all of CHs are equipped with a single antenna, and hence the transmission of each encoded packet is performed via M orthogonal time slots, follows a time-division multiple access (TDMA) schedule.In the considered network, the eavesdropper (E) attempts to obtain the source data illegally.Indeed, it tries to obtain the encoded packets to recover the original data.Follows the decoding method of FCs, the destination and the eavesdropper have to receive at least H encoded packets, where ( ) and ε is the decoding overhead which depends on concrete code design [25]- [26].Let us denote max N as the maximum number of the encoded packets that the source can send to the destination, where ( ) , respectively, while that of CHs in the conventional LEACH protocol without using the cooperative jammer technique is P .Moreover, to obtain (1), it is assumed that CH of the th m − cluster can Security-Reliability Analysis of Multi-hop LEACH Protocol with Fountain Codes and Cooperative Jamming 3 perfectly remove the interference generated by the selected jammer [14], [23], [25].
Since the eavesdropper cannot remove the interference, the instantaneous signal-to-interference-plus-noise ratio (SINR) received by the node E at this hop is expressed as From ( 1) and ( 2), the instantaneous channel capacity of the data and eavesdropping links can be calculated, respectively by ( ) Next, if the cooperative jamming technique is not used, equations ( 3) and ( 4) can be rewritten, respectively by ( ) ( ) We note that equations ( 5) and ( 6) are obtained from equations ( 3) and ( 4) by replacing α by 1. Assume that each encoded packet can be decoded successfully by CHs and the eavesdropper if the obtained instantaneous channel capacity is higher than a predetermined threshold, denoted by th C .Otherwise, the encoded packet cannot be received successfully.Due to the decode-and-forward (DF) relaying method, the probability that one encoded packet can be successfully reached to the destination can be formulated by ( ) Next, the probability that the eavesdropper can correctly receive one encoded packet is formulated as In ( 8), ( ) is the probability that the eavesdropper can successfully receive the encoded packet at the th m − hop, and this event occurs when i) the transmission at the previous hops must be successful ( ( ) obtain the encoded packet at the previous hops ( ( ) We note that the probability that the destination and the eavesdropper cannot successfully receive one encoded packet is given by D 1 m − and E 1 m − , respectively.

Channel Model
Assume that all of the link channels are Rayleigh fading, hence the channel gains D,m γ , E, 1 m γ − and J,m γ are exponential random variables (RVs) whose cumulative distribution functions (CDFs) are given, respectively as where D, E, , m m λ λ and J,m λ are parameters of D,m γ , E,m γ and J,m γ , respectively, which can be modelled as in [27]: ( )

Outage Probability (OP) and Intercept Probability (IP)
As mentioned above, outage probability (OP) of the data link is defined as the probability that the destination cannot receive at least H encoded packets after the source transmitted max N times.Therefore, OP can be computed by Equation ( 23) implies that after the source stops the transmission, the destination only receives t encoded packets, where 0 1 t H ≤ ≤ − .Let us consider the intercept probability (IP), which can be calculated as

( ) (
) In (24), since the eavesdropper can obtain w encoded packets, where max , H w N ≤ ≤ it can recover the original data of the source, and hence the source data is intercepted.

Simulation Results
In this section, we present Monte-Carlo simulations to verify the theoretical results shown in Section 3 as well as to compare the performance of the proposed protocol (denoted by Jam) and that of the conventional LEACH protocol (denoted by Non).In the simulation environment, a two-dimensional Oxy plane is considered, where all of the nodes in the th m − cluster is located at ( / ,0 m M ) and the eavesdropper is placed at (0.5, 0.5), where 0,1,..., m M = . In all of the simulations, we fix the path-loss exponent ( β ) by 3, and the target rate ( th C ) by 1.
In Figs.2-3, we present OP and IP as a function of the transmit SNR ( ) D in dB.In these figures, the fraction α is fixed by 0.85, and the values of H and max N are assigned by 5 and 7, respectively.In Fig. 2, we see that the OP values decrease with the increasing of D .Moreover, OP of the proposed protocol (Jam) is higher than that of the non-jamming protocol (Non) because the transmit power of CHs in the non-jam protocol is higher.As we can see, the performance loss is about 1dB.It is also seen from Fig. 2  considered protocols is worse with higher number of hops.In Fig. 3, the IP values rapidly increase with the increasing of D , and IP of the proposed protocol is much lower than that of the Non protocol.In addition, the IP values decrease with higher number of hops between the source and the destination (M).Finally, it is worth noting that the simulation results (Sim) match very well with the theoretical ones (Theory), which validates the formulas derived in Section 3.  N As we can see in Fig. 6, the OP performance of the considered protocols is better as H decreases and max N increases.
Indeed, with lower value of H and higher value of max N , the probability that the destination can receive sufficient number of the encoded packets increases, which reduces the outage probability.However, as illustrated in Fig. 7, the IP values also increase as H decreases and max N increases.Therefore, the values of H and max N should be designed appropriately to guarantee QoS (OP) and security (IP).For example, in the proposed protocol, to

Conclusion
In this paper, we proposed the FCs-based multi-hop LEACH protocol using cooperative jamming technique for the reliable and secure communication in WSNs.The system performance such as OP and IP was evaluated via both Monte Carlo simulations and theoretical analyzes.The obtained results showed that there exists a trade-off between security and reliability.Therefore, the system parameters such as the number of hops, the fraction of the transmit power allocated for generating the artificial noises, the number of encoded packets required for recovering the source data, and the maximum number of transmission times at the source should be carefully designed so that the proposed protocol can guarantee both security and QoS.

Figure 1 .
Figure 1.System model of the proposed protocol.
max N H ≥ .This means that after sending max N encoded packets, the source will stop the transmission.If the destination cannot successfully receive at least H encoded packets, it cannot recover the original data, which refers to an outage event.Otherwise, the data transmission between the source and the destination is successful.For the eavesdropper, if it can receive at least H encoded packets, the source data is intercepted.Let us denote D,m γ as the channel gain between CH of the ( ) 1 th m − − cluster and CH of the th m cluster and the eavesdropper, and J,m γ as the channel gain between the selected jammer of the th m − cluster and the eavesdropper.Considering the communication at the th m − hop; the instantaneous signal-to-noise ratio (SNR) received at CH of the th m − cluster is given by is variance of additive white Gaussian noise (AWGN) at CH of the th m − cluster, P is the maximum transmit power of all of CHs, P α is the transmit power of CHs in the proposed protocol ( ) transmit SNR.It is worth noting that for a fair comparison, the transmit powers of CHs and the selected jammers in the proposed protocol are P α and cluster and CH of the th m − cluster, between CH of the ( ) 1 th m − − cluster and the eavesdropper, and between the chosen jammer of the th m − cluster and the eavesdropper, respectively.From (9), the corresponding probability density functions (PDFs) are written by on Industrial Networks and Intelligent Systems 12 2018 -03 2019 | Volume 6 | Issue 18| e2 that the OP performance of the EAI Endorsed Transactions on Industrial Networks and Intelligent Systems 12 2018 -03 2019 | Volume 6 | Issue 18| e2

Figure 4 .Figure 5 .Figures 6
Figure 4. OP as a function of M when 20 dB, D = 5, H = max 6. N = satisfy the required QoS of OP 0.01 ≤ and the security level of IP 0.1 ≤ , we have to set the values of H and 9, respectively.Finally, we again see that the simulation and theoretical results are in a good agreement.