International Journal of Sensors Wireless Communications and Control - Volume 9, Issue 1, 2019

Over the decades, wireless sensor networks (WSNs) have reached its greatest heights and started to emerge into various applications, ranging from health care to multimedia transmission. In these application domains, smart autonomous low power tiny devices known as sensor nodes form a wireless network to transmit their sensed data to the base station (BS) via multi-hop routing or directly. Implementation of routing techniques in WSNs is significantly challenging due to the resourceconstrained nature of the sensor nodes. Therefore, WSN researchers have turned to different Computational intelligence (CI) techniques in an attempt to design efficient routing protocols in WSN. However, the conventional routing protocols based on computational intelligence techniques have some drawbacks viz., slow convergence rate, large memory constraints, highly sensitive to initial value, large communication overheads, and high learning period. These issues have received considerable research attention at the network layer, which leads to the development of hybrid intelligence techniques to address the routing problems. Therefore, this paper presents a systematic survey on hybrid intelligence techniques based routing protocols in WSNs. Moreover, a comparative analysis of reviewed protocols with their strengths and limitations is also included in the survey.

Background & Objective: This paper proposes a Fourier transform inspired method to classify human activities from time series sensor data. Methods: Our method begins by decomposing 1D input signal into 2D patterns, which is motivated by the Fourier conversion. The decomposition is helped by Long Short-Term Memory (LSTM) which captures the temporal dependency from the signal and then produces encoded sequences. The sequences, once arranged into the 2D array, can represent the fingerprints of the signals. The benefit of such transformation is that we can exploit the recent advances of the deep learning models for the image classification such as Convolutional Neural Network (CNN). Results: The proposed model, as a result, is the combination of LSTM and CNN. We evaluate the model over two data sets. For the first data set, which is more standardized than the other, our model outperforms previous works or at least equal. In the case of the second data set, we devise the schemes to generate training and testing data by changing the parameters of the window size, the sliding size, and the labeling scheme. Conclusion: The evaluation results show that the accuracy is over 95% for some cases. We also analyze the effect of the parameters on the performance.

Background & Objective: Multimedia aggregates various types of media such as audio, video, images, animations, etc., to form a rich media content which produces an everlasting effect in the minds of the people. Methods: In order to process multimedia applications using mobile devices, we encounter a big challenge as these devices have limited resources and power. To address these limitations, in this work, we have proposed an efficient approach named as mMedia, wherein multimedia applications will utilize the multi cloud environment using Mobile Cloud Computing (MCC), for faster processing. The proposed approach selects the best available network. The authors have also considered using the Lyapunov optimization technique for efficient transmission between the mobile device and the cloud. Results: The simulation results indicate that mMedia can be useful for various multimedia applications by considering the energy delay tradeoff decision. Conclusion: The results have been compared alongside the base algorithm SALSA on the basis of different parameters like time average queue backlog, delay and time average utility and indicate that the mMedia outperforms in all the aspects.

Background & Objective: Wireless Sensor Network (WSN) consist of huge number of tiny senor nodes. WSN collects environmental data and sends to the base station through multi-hop wireless communication. QoS is the salient aspect in wireless sensor networks that satisfies end-to-end QoS requirement on different parameters such as energy, network lifetime, packets delivery ratio and delay. Among them Energy consumption is the most important and challenging factor in WSN, since the senor nodes are made by battery reserved that tends towards life time of sensor networks. Methods: In this work an Improve-Energy Aware Multi-hop Multi-path Hierarchy (I-EAMMH) QoS based routing approach has been proposed and evaluated that reduces energy consumption and delivers data packets within time by selecting optimum cost path among discovered routes which extends network life time. Results & Conclusion: Simulation has been done in MATLAB on varying number of rounds 400- 2000 to checked the performance of proposed approach. I-EAMMH is compared with existing routing protocols namely EAMMH and LEACH and performs better in terms of end-to-end-delay, packet delivery ratio, as well as reduces the energy consumption 13%-19% and prolongs network lifetime 9%- 14%.

Background & Objective: Wireless Sensor Networks (WSN) are one of the most important elements in the Internet of Things (IoT) paradigm. It is envisaged that WSNs will seamlessly bridge the physical world with the Internet resulting in countless IoT applications in smart cities, wearable devices, smart grids, smart retails amongst others. It is necessary, however, to consider that sensing, processing and communicating large amounts of sensor data is an energy-demanding tasks. Recharging or replacing those battery-powered sensor nodes deployed in inaccessible locations is generally a tedious and time-consuming task. As a result, energy efficient approaches for WSN need to be devised in order to prolong the longevity of the network. Methods: In this paper, we present an approach that reduces energy consumption by controlling the sampling rate and the number of actively communicating nodes. The proposed approach applies compressive sensing to reduce the sampling rate and a statistical approach to decrease the sample size of sensor nodes. Results and Conclusion: The proposed approach is expected to significantly increase the lifetime of the network whilst maintaining the event detection accuracy.

Background & Objective: Vehicular Ad hoc Network (VANET) now-a-days is popular research area due to numerous challenges such as congestion control, network management, and security issues. Methods: VANET is a subtype of Mobile Ad hoc Network (MANET), therefore, allows a vehicle to exchange information with neighboring vehicles and with roadside infrastructure units. High mobility of vehicles results to frequent link breakages. The existing VANET routing protocols follow a trivial link repair mechanism. Therefore, there is a need to compute stable and reliable routes for vehicular communication as well as frequently repair the broken links. Ad hoc On-demand Distance Vector (AODV) routing protocol is a reactive routing protocol which is suitable for very dynamic environment such as VANET. In this paper, an improved Randomized Link Repair based AODV (RLRAODV) routing protocol is proposed which uses the multiple route reply packets for data transmission through alternate paths during link failure. The proposed methodology also reduces the network congestion overhead generated due to broadcasting of route request packets. Results & Conclusion: The simulation analysis shows that RLR-AODV gives better performance with respect to performance parameters such as average end-to-end delay, average packet delivery ratio, and average packet loss ratio as compared to AODV and AOMDV.

Background and Objective: Two different variants of secure routing algorithms are proposed in the present work. In both the variants a stable route is established between source and destination. Methods: The selected route is associated with the nodes having sufficient energy to establish the route and to transmit the data packets, minimum velocity to reduce the frequency of link failure, maximum distance from the source node to reduce the number of hops in the selected route, minimum number of neighbors to reduce the routing overhead. Each variant has two phases. In the first phase it is assumed that the selected route has no attackers. The first phase is made more realistic in the second phase by considering the presence of attacker in the selected route. In the first variant a node associated with the selected route overhears the transmission of the next forwarder node to detect an attacker node in the routing path. But a node may fail to overhear its next hop in presence of hidden node, due to limited overhear range etc. Such problem is eliminated in the second variant. In the second variant each node associated with the selected route searches its data packet buffer for the reception of the next data packet from its predecessor node associated with the same route and suspects the predecessor node as an attacker in case the next data packet is not found in the buffer. The performance of both the variants is compared on the basis of packet delivery ratio, throughput and average end-to-end delay. Results: The throughput and packet delivery ratio are higher in the second variant than the first variant whereas the average end-to-end delay is less in the second variant than the first variant. Conclusion: Moreover both the variants outperform the existing schemes in terms of packet delivery ratio, throughput and average end-to-end delay.

Background & Objective: In this research work, the basic digital communication system with frequency shift keying modulation technique has been implemented in MATLAB. Methods: The same sinusoidal signal in audio frequency range has been transmitted and reconstructed with the different number of quantization level and sampling rate, where sampling frequency has been taken according to the Nyquist theorem. It has been seen that reconstructed signal at low data rate has appeared with better resolution compared with the reconstructed signal at the high data rate. Results and Conclusion: In this regard, it is evident that the resolution of a reconstructed signal not only depends on data rate but it also depends on the ratio of sampling rate to quantization level. So it is required to optimize the ratio of sampling rate to quantization level before transmitting audio signal.

Background and Objective: The inter-vehicle communication is a potential issue for improving road safety, detecting traffic accidents, etc. in vehicular ad hoc network. The communication among the vehicles in VANET must be protected from the dissemination of message by unauthorized vehicles and the alteration of the message by misbehaving vehicles. Methods: In this paper, a low overhead digital watermark based vehicle revocation scheme is proposed. The sender vehicle generates a message and a random number after observing an event. It generates a deformed version of the generated message and message digest, and concatenates them. The watermark bits are generated by the sender vehicle from its unique identification and the random number. The sender vehicle embeds the watermark bits in the concatenated form of the message digest and deformed version of the generated message, and broadcasts the embedded message for its neighbours. The neighbour vehicles extract the required information from the received embedded message to verify the authentication of the sender vehicle and to identify whether the sender vehicle is an alteration attacker. It revokes unauthentic vehicles and vehicles that are identified as alteration attacker from vehicular ad-hoc network without any dependency on the trusted third party. The cracking probability and cracking time are used to measure the robustness of the scheme. The cracking probability and cracking time are measured to set the design guideline regarding the size of the watermark. The qualitative performance of the scheme is measured in terms of storage, communication and computation overhead. The significant reduction of all such overheads is observed by comparing the qualitative performance of the proposed scheme with two existing schemes. Conclusion and Results: Thus, the proposed scheme is a low overhead solution of securing vehicular ad hoc network. Performance of the scheme is also studied quantitatively in terms of the time of verifying vehicle authentication at the receiving end, delay in message dissemination at the sending end and delay in message reception at the receiving end. The quantitative performance is also compared with two of the existing schemes.

Background & Objective: The most important issue concerning the security of biometric authentication systems is protection of biometric templates. This is because once the biometric template being attacked, it cannot be canceled and reissued. Thus, the intruder could avail the facilities that are meant only for the genuine user just bypassing the enrollment phase. Methods: To protect the biometric templates from attacks, it is advantageous to modify them before storing in the databases through some cancelable and non-invertible transformations. Hence, an approach based on cancelable biometrics is proposed in this paper for providing security and privacy to biometric templates. It uses left iris and right iris as input biometric traits. Different experiments are carried out to authenticate the proposed approach. Results & Conclusion: It satisfies all template protection requirements and expected to show good recognition performance without degrading it.