Sage Journals: Discover world-class research

Abstract

Source Location Privacy (SLP) in Wireless Sensor Networks (WSNs) refers to a set of techniques and strategies used to safeguard the anonymity and confidentiality of the locations of sensor nodes (SNs) that are the source of transmitted data within the network. This protection is important in different WSN application areas like environmental monitoring, surveillance, and healthcare systems, where the revelation of the accurate location of SNs can pose security and privacy risks. Therefore, this study presents metaheuristics with sequential assignment routing based false packet forwarding scheme (MSAR-FPFS) for source location privacy protection (SLPP) on WSN. The contributions of the MSAR-FPFS method revolve around enhancing SLP protection in WSNs through the introduction of dual-routing, SAR technique with phantom nodes (PNs), and an optimization algorithm. In the presented MSAR-FPFS method, PNs are used for the rotation of dummy packets using the SAR technique, which helps to prevent the adversary from original data transmission. Next, the MSAR-FPFS technique uses an improved reptile search algorithm (IRSA) for the optimal selection of routes for real packet transmission. Moreover, the IRSA technique computes a fitness function (FF) comprising three parameters namely residual energy (RE), distance to BS (DBS), and node degree (ND). The experimental evaluation of the MSAR-FPFS system was investigated under different factors and the outputs show the promising achievement of the MSAR-FPFS system compared to other existing models.

Keywords

Wireless sensor networks metaheuristics source location privacy preserving fitness function routing false data forwarding

1 Introduction

The Internet of Things (IoT) is an integral technology of Industry 4.0, in which many smart gadgets or devices are connected to communication networks [1]. The gadgets process the outgoing and incoming data ubiquitously, and they independently interact with each other. Hence, IoT remains at the core of evolving new experiences and applications which include smart homes, smart industries, smart wearable devices, smart agriculture, etc. [2]. The smart gadgets of IoT are playing a significant part in forwarding, gathering, and processing data from source to sink nodes via multi-hop networks [3]. These autonomous and ubiquitous communication outcomes in WSN include innovative challenges in terms of their sustainability. From one perspective, as these devices have limited capabilities such as range, energy supply, computing power, transmission, etc., more and more research works have mainly focused on developing routing techniques that deal with these devices [4]. Instead, such networks are mostly operated and deployed in unattended environments. So, such networks are perhaps prone to several kinds of attacks that may affect the functioning of the network completely or partially, like finding the position of critical network nodes (sink nodes and SNs) [5]. Securing the position of the critical node in WSN becomes a crucial task that should be handled on various fronts.

Protecting source location privacy (SLP) has been gaining more significance in various application fields, specifically in security-critical circumstances [6]. The core ideology is to delay the hacker in finding source locations with the use of SLP routing protocols. In a non-SLP-protected network, a feeble aggressor, for example, a dispersed eavesdropping invader may backtrack along the message path via the network to capture the asset and identify the source location [7]. Thus, to overcome this SLP issue, different classes of SLP-aware routing techniques were devised. The phantom routing was devised for safeguarding the SLP by making some changes in the flooding routing protocol that contains a flooding and initial directed random walk. Even though the random walk-related techniques defend the SLP in theory [8], it is met with the problem of link failure and the message could not be sent to the sink (low data delivery) successfully. The method of fake sources is put on a sensor node (SN) in the network since fake sources transfer fake messages, so an opponent could not able to find which is the real one [9]. These class methods can protect the SLP effectively costing higher energy utilization of SN. Another method in delay-tolerant networks adopted the messages delay method to reach near-optimum SLP [10], costing delivery latency. Hence, many existing routing protocols have incurred high overheads (spatial or temporal) to offer SLP.

This study focuses on the design of metaheuristics with a sequential assignment routing-based false packet forwarding scheme (MSAR-FPFS) for source location privacy protection (SLPP) on WSN. The presented MSAR-FPFS model performs false packet and true packet forwarding via two separate routes to accomplish SLPP. In addition, the presented MSAR-FPFS model uses the SAR technique for false packet forwarding by phantom nodes (PNs). Furthermore, the MSAR-FPFS technique uses an improved reptile search algorithm (IRSA) for the optimal selection of routes for real packet transmission. The performance validation of the MSAR-FPFS technique is investigated under different factors. In short, the key contributions of the study are summarized as follows.

Develop an MSAR-FPFS model which comprises a dual-route forwarding mechanism, where false packets and true packets are forwarded separately. This innovative approach enhances source location privacy protection by introducing confusion for potential adversaries.

Employ the SAR approach, utilizing phantom nodes (PNs) to facilitate the forwarding of false packets. This technique plays a crucial role in obfuscating the origin of genuine data transmission.

Design an optimal Route Selection with IRSA for the optimal selection of routes for real packet transmission. By considering factors such as residual energy (RE), distance to the base station (DBS), and node degree (ND), the algorithm ensures efficient and effective data routing.

The rest of the paper is organized as follows. Section 2 provides the related works and section 3 offers the system model. Next, section 4 discusses the proposed model. Then, section 5 gives the result analysis and section 6 concludes the paper.

2 Related Works

Shukla et al. [11] present an innovative technique called SLP-RRRCT to guarantee SN location privacy through the SLPP method by randomized ring routing and confounding broadcast. The SLP-RRRCT can be benefitted from random routing including confusing transmission by making a distraction of the adversary backtracking procedure. As soon as packets were sent to the base station (BS) by neighbour grid-based ring routing once the random routing phase expired. In [12], a Jellyfish Dynamic Routing Protocol (JDRP) for conserving congestion avoidance and position confidentiality by having minimum delay guaranteed was devised. The whole sensor field can be classified as various subdivisions with this routing approach and all subdivisions opt for a targeted zone by calculating its communication distance. The dynamic routing protocol contains a virtual ring named a radial line and bell nodes termed tentacle nodes use a lot of nodes to build the network.

In [13], three source privacy protection techniques depend on clustering techniques for protecting contextual privacy. Such techniques are a hybrid method, dynamic tree and shortest path methods. Remarkably, to split the network into various square clusters, a grid-based clustering method was implemented. Zhou and Shan [14] modelled a new method based on a random walk called the multi-branch SLPP approach. This technique masks the position of real SNs by setting many proxy sources. Furthermore, the proxy source was selective arbitrarily, thwarting attackers from gaining a position-associated dataset of real SN.

In [15], a protection structure related to sector phantom routing (PSSPR) routing to solve SLP problems. The authors used the coordinates of centre node V for dividing the sector domain which plays a significant part in generating novel PNs. The PN executes some routing policy to guarantee that it selects various locations. In [16], an innovative privacy protection method called position non-dependent and section-based SLP was advanced. A greedy walk and biased random walk utilizing a 3 or 4-stage routing strategy were used here, where various stages depended on the network segment where the source was situated.

In [17], the authors proposed a probabilistic SLP (PSLP) protection method for WSN and focused on the SLP issue in WSN. To predict the state of the source, a robust adversary has used a hidden Markov model (HMM). To deal with this type of adversary, fake sources and PNs were accountable for mimicking the source behaviour employed for diversifying the routing path. In [18], a technique that depends on the cloud utilizing multi-sinks (CPSLP) was modelled for solving the problem of SLP. The authors modelled a technique that changes packet destination arbitrarily in all transmissions. As well, many sinks were implemented to create several routing paths. The outline of the intermediate node presents the routing path more flexible and random.

3 System Model

3.1 Network model

A standard Panda-Hunter method is used to study the security of SLP in this paper. Usually, pandas are supposed to inhabit a surveillance zone with a considerable amount of uniformly and randomly distributed SNs [15]. The SN becomes the SN once the panda is detected and frequently stated monitored messages to sink nodes in a multi-hop manner. The hunter is capable of acquiring the immediate place of SNs in the way of a backtrack packet by considering the communication signal. Once the hunter entered the visible region of the source position, the exposed source position. The following assumptions we make for the network: 1) All SNs in the network are deployed uniformly and randomly in the sensory region. 2) The sink node was considered the last target of the data packet and remained in the whole network centre. After deployment, the position of every SN remains the same. 3) Any two SNs in the network communicate via more than one hop. Furthermore, every SN has similar features that imply they have cache memory, similar computation capability, and initial energy.

3.2 Attack model

Let us patient attacker mode with smaller network visibility in the standard SLP protection scenario. The adversary only can eavesdrop on any data for the packet in the sensor network and cannot interfere with the normal operation of sensory networks. According to a certain routing protocol, the information monitored by the sources is often transferred to the sink destination node. Thus, under a normal scenario, once the attacker is primarily located at the sink node, then it guarantees to effectively capture every individual packet. The adversary is an illegitimate attacker based on huge profits, viz., equipped with effective tools to eavesdrop on some data in the source. Once the aggressor overhears the novel message, it measures the arriving angle and the signal strength which is received to recognize the immediate transmitter node; it won’t have some powerful effect on the network. Afterwards, the attacker implements passive attacks by initiating back tracing packet path by moving to the prior nodes near the source till the SN is reached.

4 The Proposed Method

In this article, we have developed a novel MSAR-FPFS approach to achieve SLPP on the WSN. The presented MSAR-FPFS technique follows a two-stage process: false packet forwarding and real packet forwarding. Initially, the MSAR-FPFS technique enables the PNs to forward false packets in the WSN using the routes discovered by the SAR technique. At the next stage, the real packets are forwarded by the IRSA technique to accomplish privacy-preserving data transmission in the network. At the same time, the IRSA technique computes an FF comprising three parameters they are ND, RE, and DBS. Figure 1 illustrates the overall procedure of the MSAR-FPFS system.

Fig. 1

Overall procedure of MSAR-FPFS approach.

4.1 Design of SAR-based False Packet Forwarding

In this work, the false packets can be forwarded by the PNs using the SAR technique to BS. PN has been placed on every corner grid of the network [19]. Also, the source of fake data packets from PN can be rotated in a round-robin manner for transmitting the original data packets in the SNs to BS. The aggressor believes that the SN continuously changes its position. At last, the original data packets were transferred at the BS through neighbour gird-based ring routing. It is shown that the distance and count of hops from SNs to BS are different for all the routing data packets, and also all data packets from SNs to the BS follow a random path.

In the MSAR-FPFS approach, the adversary might be misled by the stimulated packet communication technique through the PNs. Once the adversary has received the PN’s dummy packet and backtracked the routing path, the SN’s safety hours can be improved. The fake packet communication from the PN in a round-robin manner might increase the privacy of SNs and generate a dilemma in the aggressor’s mind regarding the movement of SNs. In the MSAR-FPFS method, the stimulated packet communication technique is used, however, the count of hops in stimulated packet routing was constrained. Therefore, the presented technique raised the privacy of SNs and depleted non-hot spot area energy.

The SAR technique constructs various trees, where the root of every tree is a single-hop neighbour of the sink. Every tree grows outwards from the sink and ignores nodes having higher delay or lower throughput. Most of the node belongs to multiple trees at the end of the procedure. The tree rooted at A & B, the single-hop neighbour of the sink, is demonstrated. Node C belongs to these two trees, and correspondingly has path lengths of 3 & 5, to the sink, using the two trees. Every SN records the two parameters regarded every path through it: the additive quality of services (QoS) metrics like delay and the accessible energy resource on the path. This enables the node to select a single path from among multiple relays for its data to sink. The SAR technique selects a path with higher assessed energy sources, and provision is made for accommodating the packet of various priorities. A weighted QoS metric handles prioritized packet that is calculated by the product of delay and priority level. The routing makes sure that the similar weight QoS metric can be preserved. Therefore, low priority packet uses the path of high delay, and high priority packet takes low delay path. For instance, if node C creates the data packet of priority 3, which follows the long path alongside tree B, and the data packet of priority 5 (highest priority) follows the short path alongside tree A such that the priority °ø delay QoS metrics are preserved. SAR diminishes the average weight of QoS metrics over the network lifespan. The sink sporadically triggers a metric update to reflect changes in accessible energy resources afterward a few communications.

4.2 Design of IRSA-based Real Packet Forwarding

The IRSA technique computes a FF comprising three parameters namely RE, DBS, and ND. The hunting and the encircling stages are the two different phases of RSA [20]. The RSA shifts among the hunting search phase and the encircling phase and the switching among various phases can be done by separating the iteration count into four diverse segments. RSA commences with an initially generated set of candidate solutions stochastically utilizing Equation (1):

$χ_{jk} = rand \times (U_{b} - L_{b}) + L_{b} k = 1, 2, \dots n$ (1)

χ_jk =Initialization matrix, j = 1, 2….P, where rand is a randomly generated value in the range [0, 1], L_b, and U_b, designate the lower and upper bounds, n characterizes dimensions (columns of the initialized matrix) of the problem and P indicates population size (rows of initialized matrix). Figure 2 defines the steps involved in IRSA.

Fig. 2

Steps involved in IRSA.

The encircling phase was an exploration of the high-density area. High walking and belly walking were the two phases of the encircling stage based on the movements of the crocodile which plays a major role. This movement helps to discover a wide search space however it doesn’t help to catch the prey.

$\begin{matrix} χ_{jk (τ + 1) = {Best}_{k} (τ)} \\ \times (- μ_{(jk)} (τ)) \times β - (R_{(jk)} (τ) \times rand), τ ⩽ \frac{T}{4} \end{matrix}$ (2)

$\begin{matrix} x_{jk} (τ + 1) = {Best}_{k} (τ) \times χ_{(r_{1}, k)} \\ \times ES (τ) \times rand, τ ⩽ 2 \frac{T}{4} and τ > \frac{T}{4} \end{matrix}$ (3)

Let Best_k (τ) be the optimum solution at the kth location, rand signifies a random value, τ denotes the existing iteration number, and the maximal amount of iterations can be denoted as T. μ_(j,k) denotes hunting operator value of the jth solution at the kth location. The μ_(j,k) value is defined by:

$μ_{(j, k)} = {Best}_{k} (τ) \times P_{(j, k)}$ (4)

In Equation (4), β indicates the sensitivity parameter, and it defines exploration accuracy. Another intention of the R_(j,k) function is to lessen the search range that is calculated by Equation (5):

$R_{(j, k)} = \frac{{Best}_{k} (τ) - P_{(r_{2^{'}} k)}}{{Best}_{k} (τ) + e}$ (5)

In Equation (5), r₁ indicates the value of a random integer which ranges from 1 and N. Now, N signifies the overall amount of candidate solutions. z_(r1,l) denotes an arbitrary location for the kth solution. r₂ was a random (arbitrary) number ranging from 1 and N, whereas e characterizes a smaller magnitude value. Evolutionary Sense (ES) denoted by τ was a probability-based ratio and the scientific expression of ES can be represented below:

$ES (τ) = 2 \times r_{3} \times (1 - \frac{1}{T})$ (6)

In Equation (6), r₃ characterizes a random integer. P_(j,k) can be evaluated as follows:

$P_{(j, k)} = α + \frac{χ_{(j, k) - M (x_{j})}}{{Best}_{k} (τ) x (U_{b (k)} - (L_{b (k)}) + e}$ (7)

In Equation (7), α denotes a sensitivity limit which controls the exploration accuracy M (χ_j) represents the average position of the j_th solution and it is analyzed by Equation (8):

$M (χ_{j) = \frac{1}{n}} \sum_{k = 1}^{n} χ_{(j, k)}$ (8)

In resemblance to the encircling stage, the hunting stage consists of different policies like hunting cooperation and coordination. For the exploitation phase, Equations (9) and (10) are used:

$\begin{matrix} x (j, k) (τ + 1) = {Best}_{k} (τ) \times (P_{(j, k)} \\ (τ)) \times rand, τ ⩽ 3 \frac{T}{4} and τ > 2 \frac{T}{4} \end{matrix}$ (9)

$\begin{matrix} x (j, k) (τ + 1) = {Best}_{k} (τ) - μ_{(j, k)} (τ) \\ \times e - R_{(j, k)} (τ) \times rand, τ ⩽ Tand τ > 3 \frac{T}{4} \end{matrix}$ (10)

Where Best_k (τ) represents the kth position in the optimum solution at the present iteration. Likewise, μ_(j,k) denotes the hunting operator that can be evaluated by Equation (4).

The RSA is a recently established optimizer method that could effectively resolve different optimization problems. However, the RSA possess certain shortcomings while resolving high dimensional nonconvex optimization problems like local minima trapping, slower convergence speed and high computational complexity. Thus, certain modifications are proposed to the original RSA technique to overcome these issues. To avoid trapping local minima it needs the solution candidate for widely exploring the search space. Thus, to improve exploration, a sine operator was added in the RSA technique’s higher walking phase. These adjustments have been stimulated by the dynamic exploration method in the sine cosine model. The sine operator offers a global exploration ability. In IRSA, with the sine operator, Equation (2) was replaced with the subsequent formula.

$\begin{matrix} χ_{jk} (τ + 1) = {Best}_{k} (τ) + (r_{1} \times \sin (rand) \\ \times | r_{2} \times {Best}_{k} (τ) - xjk |, for τ ⩽ \frac{T}{3} \end{matrix}$ (11)

In Equation (11), r₁, r₂, and rand are randomly selected values ranging from zero to one. χ_jk indicates the existing location, and Best_k indicates a better solution. The Levy flight was a random procedure which follows the Levy dispersal function:

$levy = 0.01 \times \frac{u}{v^{\frac{1}{ζ}}}$ (12)

In Equation (12), u and v obey uniform distribution.

$u \sim (0, σ_{u}^{2}), v \sim (0, σ_{v}^{2})$ (13)

$σ_{u} = (\frac{δ (1 + ζ) \sin \frac{π ζ}{2}}{δ [\frac{1 + ζ}{2}] ζ * 2^{ζ - \frac{1}{2}}})^{\frac{1}{ζ}}$ (14)

$σ_{v} = 1$ (15)

From the expression, δ denotes the average gamma function. ζ indicates the curial parameter in Levy fight that defines the jump size. The lesser value of ζ leads to a smaller arbitrary step. This allows the solution candidate to find the region near to attained solution that enhances exploitation abilities. This exploitation assurances global convergence. Thus, rather than Equation (10) the Levy operator was employed for updating the location in the concluding phases of IRSA

$\begin{matrix} x_{jk} (τ + 1) = {Best}_{k} (τ) + randn \times levy \oplus \\ (x_{jk} - {Best}_{k} (τ)), for τ ⩽ Tand τ > 3 \frac{T}{4} \end{matrix}$ (16)

Equation (16), ⊕ denotes the entry-wise multiplication, and randn represents a uniform distribution random integer. This improvisation considerably decreases the difficulty of the model. In the primary RSA, Equations (4) and (5) were extremely complicated and increased the time difficulty of the method. But, we can eliminate the equation from the algorithm with the abovementioned adaptation. This implies that the presented IRSA algorithm doesn’t have to calculate the equation that leads to a nearly 3-to-4-fold lessening in time complication. High efficiency, higher global exploration, low time complexity, and high convergence speed are the improvements observed in the presented method. Algorithm 1 represents the IRSA’s pseudocode.

Algorithm 1: IRSA Pseudocode
Initializing random population x
Initializing iteration counter τ = 0, maximal iteration T, alpha, beta while τ < T
Assess the fitness of possible candidate
Defines the optimum solution
Upgrade Es, P_(j,k) by utilizing Equations (6) and (7)
For j = 1: p
for k = 1 : n
If $τ ⩽ \frac{T}{3}$
Resolve using Equation (11)
Else if $τ ⩽ 2_{4}^{T}$ and τ > 3T
Resolve using Equation (3)
Else if $τ ⩽ 3_{4}^{T}$ and $τ > 2 \frac{T}{4}$
Resolve using Equation (9)
Else
Resolve utilizing Equation (16)
End if
End for
End for
t = t + 1
End while
Return optimum solution

Based on the following parameters, the FF can be derived.

RE: The objective in mapping every cluster head (CH) to the Next-Hop node (NHN) based on the remaining energy of NHN [21]. A CH node chooses a node amongst potential NHN so that it has additional RE when compared to others.

Objective 1:

$Maximize : h_{1} = \sum_{j = 1}^{m} E_{{CH}_{j}}$ (17)

DBS: The objective in mapping all CH to NHN is based on the distance between NHN and BS. A CH node chooses CH amongst potential NHN so that it has reduced distance between the BS and other potential nodes.

$\begin{matrix} Minimize : h_{2} = \sum_{j = 1}^{m} dis ({CH}_{j}, NH ({CH}_{j})) \\ + dis (NH ({CH}_{j}) + BS) \end{matrix}$ (18)

ND: Each CH mapping to the NHNs is based on the ND of the NHNs. A CH is assigned to the NHN evaluated by the minimum ND.

$Minimize : h_{3} = \sum_{j = 1}^{m} N_{d} (NH ({CH}_{j}))$ (19)

In Equation (19), a weighted combination method is implemented for lessening the objective as this doesn’t contradict one another. Hence, the following FF is implemented:

$Minimize Fitness = β_{1} \times \frac{1}{h_{1}} + β_{2} \times h_{2} + β_{3} \times h_{3}$ (20)

Where,

$0 < β_{1}, β_{2} & β_{3} < 1$ (21)

5 Results and discussion

In this segment, the investigational validation of the MSAR-FPFS technique is investigated in detail. The proposed MSAR-FPFS technique is simulated by utilizing the MATLAB tool. Table 1 defines the details of parameter settings. A detailed comparison study is made with existing models [19 , 23] such as baseline, source location protection protocol based on dynamic routing (SLPDR) [22], source location privacy protection scheme based on ring-loop routing (SLPRR) [23], and source location privacy protection scheme based on random rings and limited hop fake packet routing scheme (SLP-RRFPR) [19] techniques. In this work, the first-order radio energy model is used to analyze the energy consumption [24].

Table 1
Parameter settings

Parameters Values

SN number 800

Network area (M×M) 200×200 m²

E₀ (Deployed node’s original energy) 0.5 J

E_elec for the total nodes 50 nJ/bit

∈_fs denotes the energy depletion in the free space model 10 pJ/bit/m²

∈_mp indicates the energy depletion in the multi-path model 0.0013 pJ/bit/m⁴

L (number of bits in a packet) 4000 bits

SN transmission range 40 m

Parameters	Values
SN number	800
Network area (M×M)	200×200 m²
E₀ (Deployed node’s original energy)	0.5 J
E_elec for the total nodes	50 nJ/bit
∈_fs denotes the energy depletion in the free space model	10 pJ/bit/m²
∈_mp indicates the energy depletion in the multi-path model	0.0013 pJ/bit/m⁴
L (number of bits in a packet)	4000 bits
SN transmission range	40 m

Table 2 and Figs. 3-4 illustrates the comparative analysis of the consumed energy in the data receiving and forwarding method from the SN to BS. The distance between an SN, BS, and communication radius was different for analyzing the effect of energy consumption (ECON). Figures 3 and 4 demonstrate that flooding-based systems: probabilistic, phantom, and baseline utilize major energy because of packet communication in all the SNs from the network. The SLPDR and SLPRR have depleted the minimal energy in every comparative protocol. The presented SLP-RRFPR has depleted somewhat a little energy compared to the SLPDR and SLPRR approaches. During the presented method, it exploits the dummy packet communication with the confounding communication and with the PNs to make a distraction for the attacker’s backtracking procedures. Besides, the dummy packet communication’s count is superior to the recent algorithms. However the dummy packets were forwarded in the SNs of the hot spot field, and the SN privacy was enhanced. Thus, a further ECON is suitable. It is noted from the outcomes that once the distance between SN and BS was augmented, the energy exhaustion was enhanced. However, the ECON was decreased to enhance the communication radius because of some hops needed in the SN to BS.

Table 2

ECON analysis of MSAR-FPFS approach with other systems

Energy Consumption (J)
	Baseline	SLPDR Protocol	SLPRR Protocol	SLP-RRFPR Protocol	MSAR-FPFS
Distance Between SN and BS
20	393	10	6	21	18
40	394	16	8	21	18
60	393	16	11	21	17
80	393	15	8	22	19
100	394	18	7	23	21
120	394	15	11	24	22
140	395	21	11	23	21
160	394	18	11	24	21
Sensor Node Transmission Range (meter)
20	392	11	6	32	27
25	390	11	8	21	20
30	390	11	4	16	14
35	391	12	6	19	16
40	390	9	5	16	15

Fig. 3

ECON analysis of MSAR-FPFS approach under varying distances between SN and BS.

Fig. 4

ECON analysis of MSAR-FPFS approach under varying SNTR.

In Table 3, a detailed network lifetime (NLT) investigation of the MSAR-FPFS technique with recent models is demonstrated [19]. Figure 5 represents the NLT assessment of the MSAR-FPFS technique under varying distances between SN and BS. The results deliberated that the MSAR-FPFS technique reaches higher values of NLT under all aspects. With a distance of 20 m, the MSAR-FPFS technique gains an increasing NLT of 1623 rounds while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques obtain decreasing NLT of 1136, 371, 364, and 1590 rounds respectively. Eventually, with a distance of 160 m, the MSAR-FPFS algorithm reaches a higher NLT of 1541 rounds while the baseline, SLPDR, SLPRR, and SLP-RRFPR systems gain lesser NLT of 1136, 492, 436, and 1214 rounds correspondingly.

Table 3

NLT analysis of MSAR-FPFS approach with other systems

Network Lifetime (Rounds)
	Baseline	SLPDR Protocol	SLPRR Protocol	SLP-RRFPR Protocol	MSAR-FPFS
Distance Between SN and BS
20	1136	371	364	1590	1623
40	1132	567	318	1551	1580
60	1139	459	358	1387	1551
80	1142	452	384	1482	1577
100	1139	459	469	1570	1603
120	1142	423	433	1443	1584
140	1139	390	443	1570	1613
160	1136	492	436	1214	1541
Sensor Node Transmission Range (meter)
20	1207	430	403	1647	1772
25	1164	403	365	1484	1658
30	1131	430	370	1446	1636
35	1115	272	218	1446	1653
40	1077	267	218	1419	1620

Fig. 5

NLT analysis of MSAR-FPFS approach under varying distances between SN and BS.

Figure 6 signifies the NLT analysis of the MSAR-FPFS system under varying SN transmission ranges (SNTR). The outcomes deliberated that the MSAR-FPFS method reaches greater values of NLT under all aspects. With SNTR of 20 m, the MSAR-FPFS technique achieves a maximum NLT of 1772 rounds while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques obtain reduced NLT of 1207, 430, 403, and 1647 rounds correspondingly. Finally, with SNTR of 40 m, the MSAR-FPFS methodology reaches maximal NLT of 1620 rounds while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques attain minimal NLT of 1077, 267, 218, and 1419 rounds respectively.

Fig. 6

NLT analysis of MSAR-FPFS approach under varying SNTR.

A comparison study of the MSAR-FPFS technique with recent approaches in terms of delay (DEL) is reported in Table 4. In Fig. 7, the DEL assessment of the MSAR-FPFS technique with recent models is revealed under several distances between SN and BS. The results indicate the promising DEL values of the MSAR-FPFS technique. With a distance of 20 m, the MSAR-FPFS technique provides the least DEL of 1 hop while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques accomplish increased DEL of 19, 2, 3, and 21 hops respectively. Simultaneously, with a distance of 100 m, the MSAR-FPFS algorithm provides minimal DEL of 12 hops while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques accomplish increased DEL of 39, 17, 16, and 31 hops respectively. Concurrently, with a distance of 160 m, the MSAR-FPFS system provides least DEL of 15 hops while the baseline, SLPDR, SLPRR, and SLP-RRFPR systems accomplish increased DEL of 58, 22, 32, and 40 hops correspondingly.

Table 4

Delay analysis of MSAR-FPFS approach with other systems

Delay (Hops)
	Baseline	SLPDR Protocol	SLPRR Protocol	SLP-RRFPR Protocol	MSAR-FPFS
Distance Between SN and BS
20	19	2	3	21	1
40	24	4	3	24	2
60	27	11	9	24	5
80	37	14	12	28	7
100	39	17	16	31	12
120	43	18	18	34	12
140	49	18	25	38	13
160	58	22	32	40	15
Sensor Node Transmission Range (meter)
20	32	14	6	2	1
25	30	10	5	3	1
30	27	8	5	3	1
35	25	8	4	2	1
40	21	5	4	2	1

Fig. 7

Delay analysis of MSAR-FPFS approach under varying distances between SN and BS.

In Fig. 8, the DEL investigation of the MSAR-FPFS technique with recent approaches is exposed under several SNTRs. The outcomes indicate the promising DEL values of the MSAR-FPFS system. With SNTR of 20 m, the MSAR-FPFS method gives reduced DEL of 1 hop while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques accomplish higher DEL of 32, 14, 6, and 2 hops correspondingly. At the same time, with SNTR of 30 m, the MSAR-FPFS approach provides minimal DEL of 1 hop while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques achieve enhanced DEL of 27, 8, 5, and 3 hops respectively. Simultaneously, with SNTR of 40 m, the MSAR-FPFS system provides least DEL of 1 hop while the baseline, SLPDR, SLPRR, and SLP-RRFPR algorithms realize increased DEL of 21, 5, 4, and 2 hops correspondingly.

Fig. 8

Delay analysis of MSAR-FPFS approach under varying SNTR.

In Table 5, a detailed safety period (SPD) study of the MSAR-FPFS method with recent approaches is revealed. Figure 9 exemplifies the SPD study of the MSAR-FPFS algorithm under varying distances between SN and BS. The results inferred that the MSAR-FPFS methodology obtains enhanced values of SPD under all aspects. With a distance of 20 m, the MSAR-FPFS approach attains enhanced SPD of 76 while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques obtain minimal SPD of 2, 20, 20, and 53 correspondingly. Lastly, with distance of 160 m, the MSAR-FPFS system gains increasing SPD of 154 while the baseline, SLPDR, SLPRR, and SLP-RRFPR methods achieve decreasing SPD of 6, 55, 49, and 148 correspondingly.

Table 5

Safety period analysis of MSAR-FPFS approach with other systems

Safety Period
	Baseline	SLPDR Protocol	SLPRR Protocol	SLP-RRFPR Protocol	MSAR-FPFS
Distance Between SN and BS
20	2	20	20	53	76
40	2	32	59	55	97
60	2	23	41	101	117
80	4	30	64	105	124
100	6	38	46	49	120
120	6	42	57	70	123
140	7	67	63	146	151
160	6	55	49	148	154
Sensor Node Transmission Range (meter)
20	6	51	53	60	61
25	4	41	40	48	52
30	3	30	33	43	49
35	3	22	30	32	45
40	2	23	27	26	37

Fig. 9

SPD analysis of MSAR-FPFS approach under varying distances between SN and BS.

Figure 10 indicates the SPD examination of the MSAR-FPFS system under varying SNTR. The outcomes considered that the MSAR-FPFS method reaches higher values of SPD under all aspects. With SNTR of 20 m, the MSAR-FPFS approach gains increasing SPD of 61 while the baseline, SLPDR, SLPRR, and SLP-RRFPR techniques obtain diminishing SPD of 6, 51, 53, and 60 correspondingly. At last, with SNTR of 40 m, the MSAR-FPFS technique gains maximal SPD of 37 while the baseline, SLPDR, SLPRR, and SLP-RRFPR approaches obtain minimal SPD of 2, 23, 27, and 26 correspondingly.

From these results and discussion, it is guaranteed that the MSAR-FPFS technique reaches better performance than other models.

Fig. 10

SPD analysis of MSAR-FPFS approach under varying SNTR.

6 Conclusion

In this study, we have developed a novel MSAR-FPFS approach to achieve SLPP on the WSN. The presented MSAR-FPFS technique follows a two-stage process: false packet forwarding and real packet forwarding. Initially, the MSAR-FPFS technique enables the PNs to forward false packets in the WSN using the routes discovered by the SAR technique. At the next stage, the real packets are forwarded by the IRSA technique to accomplish privacy-preserving data transmission in the network. The performance validation of the MSAR-FPFS approach is tested under different aspects and the outcomes are investigated under varying transmission ranges of SNs and distances from SNs to BS. A widespread comparison results show the promising performance of the MSAR-FPFS technique compared to other existing models. Therefore, the presented MSAR-FPFS technique can be employed for accomplishing SLPP in WSN. In future, data encryption schemes can be integrated into the MSAR-FPFS technique to improve security in WSN.

Declarations

Ethical Approval

Not Applicable

Competing interests

Authors don’t have any competing interest

Authors’ contributions

Pitchandi@Sankaralingam - Data Collection, Conceptualization

Arunachalaperumal - Data Curation, Investigation,

Arunachalaperumal –Validation, Editing

Pitchandi@Sankaralingam - Review and Editing

Funding

Not Applicable

Availability of data and materials

Data can be provided on request.

References

Roy

P.K.

, Tripathy

A.K.

, Singh

S.K.

and Li

K.C.

, Recent advancements in privacy-aware protocols of source location privacy in wireless sensor networks: a survey, Computer Science and Information Systems (00) (2022), pp. 7–7.

Alzaabi

, Aldoobi

, Alnuaimi

, Alserkal

, Alsuwaidi

, Ababneh

Grid-Based Source Location Privacy Protection Schemes in IoT Wireless Sensor Net521 works. In 2021 4th International Conference on Data Storage and Data Engineering (2021), (pp. 31–36).

Al-Ahmadi

S.A.

, Counterfeit Clones: A Novel Technique for Source and Sink Location Privacy in Wireless Sensor Networks, IEEE Access 10 (2022), 62693–62701.

Manjula

, Koduru

and Datta

, Protecting source location privacy in IoT-Enabled wireless sensor networks: the case of multiple Assets, IEEE Internet of Things Journal 9(13) (2021), 10807–10820.

Roy

P.K.

and Rimjhim

J.P.

, Singh, Finding location of fake and phantom source for source location privacy in wireless sensor network, International Journal of Communication Networks and Distributed Systems 26(1) (2021), 114–134.

Sharma

, Bhatt

Source location privacy preservation in IoT-enabled event-driven WSNs, International Journal of Pervasive Computing and Communications (ahead-of-print). 2022.

, Jhumka

and Maple

, Silence is Golden: A Source Location Privacy Scheme for Wireless Sensor Networks Based on Silent Nodes, Security and Communication Networks 2022 (2022).

Manjunath

D.R.

and Kumar

, Source location privacy for geographical routing in wireless sensor networks: SLPGR,, International Journal of Computer Networks and Applications 8(4), p. 422.

Bradbury

and Jhumka

, Quantifying Source Location Privacy Routing Performance via Divergence and Information Loss, IEEE Transactions on Information Forensics and Security 17 (2022), pp. 3890–3905.

10.

Umar

, Abetu

M.T.

, Efa

B.T.

, Dinsa

E.F.

, Prabhat

Wireless Sensor Networks Source Location privacy preservation Mechanism. In Advanced Informatics for Computing Research: 4th International Conference, ICAICR 2020, Gurugram, India, December 26–27, 2020, Revised Selected Papers, Part II (2021), (pp. 160–168). Singapore: Springer Singapore.

11.

Shukla

, Singh

, Sajwan

, Verma

and Kumar

, A source location privacy preservation scheme in WSN-assisted IoT network by randomized ring and confounding transmission,pp, Wireless Networks 28(2) (2022), 827–852.

12.

Christopher

V.B.

and Jasper

, Jellyfish dynamic routing protocol with mobile sink for location privacy and congestion avoidance in wireless sensor networks, Journal of Systems Architecture 112 (2021), 101840.

13.

Al-Mistarihi

M.F.

, Tanash

I.M.

, Yaseen

F.S.

and Darabkh

K.A.

, Protecting source location privacy in a clustered wireless sensor networks against local eavesdroppers, Mobile Networks and Applications 25 (2020), pp. 42–54.

14.

Zhou

, Shan

Multi-branch source location privacy protection scheme based on random walk in WSNs. In 2019 IEEE 4th International Conference on Cloud Computing and Big Data Analysis (ICCCBDA) (2019), (pp. 543–547). IEEE.

15.

Chen

, Sun

, Yang

, Li

, Niu

and Zhou

, PSSPR: a source location privacy protection scheme based on sector phantom routing in WSNs, International Journal of Intelligent Systems 37(2) (2022), 1204–1221.

16.

Mukamanzi

, Raja

, Koduru

, Datta

Position independent and section-based source location privacy protection in WSN, IEEE Transactions on Industrial Informatics. 2022.

17.

Wang

, Han

, Zhang

, Guizani

and Chan

, A probabilistic source location privacy protection scheme in wireless sensor networks, IEEE Transactions on Vehicular Technology 68(6) (2019), 5917–5927.

18.

Han

, Miao

, Wang

, Guizani

and Zhang

, CPSLP: A cloud-based scheme for protecting source location privacy in wireless sensor networks using multi-sinks, IEEE Transactions on Vehicular Technology 68(3) (2019), 2739–2750.

19.

Shukla

, Singh

, Sajwan

, Kumar

, Kumari

, Kumar

and Panthi

, SLP-RRFPR: a source location privacy protection scheme based on random ring and limited hop fake packet routing for wireless sensor networks, Multimedia Tools and Applications 81(8) (2022), 11145–11185.

20.

Khan

M.K.

, Zafar

M.H.

, Rashid

, Mansoor

, Moosavi

S.K.R.

and Sanfilippo

, Improved Reptile Search Optimization Algorithm: Application on Regression and Classification Problems, Applied Sciences 13(2) (2023), 945.

21.

Rao

P.S.

, Lalwani

, Banka

and Rao

G.S.N.

, Competitive swarm optimization based unequal clustering and routing algorithms (CSO-UCRA) for wireless sensor networks, Multimedia Tools and Applications 80(17) (2021), 26093–26119.

22.

Han

, Zhou

, Wang

, Zhang

and Chan

, A source location protection protocol based on dynamic routing in WSNs for the Social Internet of Things, Futur. Gener Comput Syst 82 (2018), 689–697.

23.

Wang

, Han

, Zhou

, Ansere

J.A.

and Zhang

, A Source Location privacy protection scheme based on ring-loop routing for the IoT, Comput Networks 148 (2019), 142–150.

24.

Heinzelman

W.R.

, Chandrakasan

, Balakrishnan

Energy-efficient communication protocol for wireless microsensor networks. In Proceedings of the 33rd annual Hawaii international conference on system sciences (2000), (pp. 10). IEEE.

Improved reptile search algorithm with sequential assignment routing based false packet forwarding scheme for source location privacy protection on wireless sensor networks

Abstract

Keywords

1 Introduction

2 Related Works

3 System Model

3.1 Network model

3.2 Attack model

4 The Proposed Method

4.2 Design of IRSA-based Real Packet Forwarding

Declarations

Ethical Approval

Competing interests

Authors’ contributions

Funding

Availability of data and materials

References