The winter season worldwide causes serious driving issues due to heavy fog on the roads. In situations like these, having an intelligent transportation system is crucial for safe driving. Therefore, to ensure safe and stable driving in such a natural weather situation, this paper proposes a machine learning-based software-defined vehicular network (SDVN). The SDN main controller (MC) is responsible for selecting the optimal controller (OC) from among several local controllers (LCs). The OC is accountable for sharing the load with the MC and regulating vehicular speed by disseminating alarming messages (AMs) in the form of a speed reduction, lane change, and low visibility alerts. It is also responsible for providing optimal paths from source to destination, thereby ensuring the exchange of data between vehicles in such foggy weather conditions. OC and LCs assist vehicles in calculating link visibility time (LVT) to ensure stable links in the network. To guarantee safety, the paper considers two types of vehicles: normal and stubborn (who do not follow OC’s instructions). The objective is to enable all types of vehicles to travel safely in foggy weather. To achieve this goal and evaluate its validity, the scheme considers different cases following the selection of an OC and the optimal path. Considering different cases in foggy weather, our findings indicate that vehicles can demonstrate safety and stability in low visibility conditions.

The controller selection problem (CSP) in software-defined vehicular networks (SDVNs) causes a long delay and large overhead due to sporadic links, especially when vehicles are outside the coverage of the main controller (MC) or a local controller (LC) and when distribution of the load among LCs is unequal. We solve the CSP using a support vector machine (SVM) algorithm for heterogeneous communications. The MC runs the algorithm based on vehicle density to select an optimal controller (OC) in both highway and city scenarios. After this selection, the OC is responsible for selecting a stable path to deliver packets from source to destination. This machine-learning-based SDVN scheme selects the minimum link duration ($LD$) for communication between nodes, but selects the path with the maximum path time between source and destination. The protocol has two phases: OC selection and path selection. The OC is also responsible for distributing the load equally to the other LCs. The two-phase selection scheme improves the network performance in terms of delivery ratio (maximum value of 94.2%), end-to-end delay (minimum value of 0.11 ms), and routing overhead ratio (maximum incurred is 10%), which is proved by our simulation results in comparison to an existing scheme.

Cognitive radio is becoming an increasingly important component of Internet of Things technologies, because it improves spectrum efficiency and provides a way to overcome the challenge of excessive demand for wireless communications. Research and development of cognitive radio communication systems are timely and helpful. Recently, Zheng et al. proposed a network stability-aware clustering protocol for cognitive radio sensor networks. In this letter, we discuss its fatal flaws, as well as ways to overcome them.

Due to the constantly changing sea surface, there is a high risk of link fragility caused by sea waves when different marine users are intended to establish stable links for communication. To ensure stability with less delay, finding a stable route is one of the crucial aspects of maritime networks. In order to achieve this aim, we propose a routing protocol for cognitive maritime networks based on software-defined networking (SDN). This SDN-based cognitive routing protocol provides stable routes among different marine users. To provide the global view of the whole network, a main controller is placed close to the seashore, whereas the localized views are provided by the cluster heads. Autonomous surface vehicles are used as gateways under sparse network conditions to collect and transport data among clusters, and to and from the main controller. This is an SDN-based ship-to-ship communication scheme where two ships can only establish a link when they not only have consensus about a common idle channel but are also within the communication range of each other. We perform extensive simulations to test the proposed scheme with different parameters and find better performance in comparison with both SDN-based and non-SDN-based schemes in terms of end-to-end delay, packet delivery ratio, and routing overhead ratio.

Being one of the most multifaceted cyber-physical systems, smart grids (SGs) are arguably more prone to cyber-threats. A covert data integrity assault (CDIA) on a communications network may be lethal to the reliability and safety of SG operations. They are intelligently designed to sidestep the traditional bad data detector in power control centers, and this type of assault can compromise the integrity of the data, causing a false estimation of the state that further severely distresses the entire power system operation. In this paper, we propose an unsupervised machine learning-based scheme to detect CDIAs in SG communications networks utilizing non-labeled data. The proposed scheme employs a state-of-the-art algorithm, called isolation forest, and detects CDIAs based on the hypothesis that the assault has the shortest average path length in a constructed random forest. To tackle the dimensionality issue from the growth in power systems, we use a principal component analysis-based feature extraction technique. The evaluation of the proposed scheme is carried out through standard IEEE 14-bus, 39-bus, 57-bus, and 118-bus systems. The simulation results show that the proposed scheme is proficient at handling non-labeled historical measurement datasets and results in a significant improvement in attack detection accuracy.

Cognitive radio (CR) is being considered as a vital technology to provide solution to spectrum scarcity in next generation network, by efficiently utilizing the vacant spectrum of the licensed users. Cooperative spectrum sensing in cognitive radio network has a promising performance compared to the individual sensing. However, the existence of the malicious users’ attack highly degrades the performance of the cognitive radio networks by sending falsified data also known as spectrum sensing data falsification (SSDF) to the fusion center. In this paper, we propose a double adaptive thresholding technique in order to differentiate legitimate users from doubtful and malicious users. Prior to the double adaptive approach, the maximal ratio combining (MRC) scheme is utilized to assign weight to each user such that the legitimate users experience higher weights than the malicious users. Double adaptive threshold is applied to give a fair chance to the doubtful users to ensure their credibility. A doubtful user that fails the double adaptive threshold test is declared as a malicious user. The results of the legitimate users are combined at the fusion center by utilizing Dempster-Shafer (DS) evidence theory. Effectiveness of the proposed scheme is proved through simulations by comparing with the existing schemes.

Spectrum sensing is of the utmost importance to the workings of a cognitive radio network (CRN). The spectrum has to be sensed to decide whether the cognitive radio (CR) user can transmit or not. Transmitting on unoccupied spectrum becomes a hard task if energy-constrained networks are considered. CRNs are ad hoc networks, and thus, they are energy-limited, but energy harvesting can ensure that enough energy is available for transmission, thus enabling the CRN to have a theoretically infinite lifetime. The residual energy, along with the sensing decision, determines the action in the current time slot. The transmission decision has to be grounded on the sensing outcome, and thus, a combined sensing–transmission framework for the CRN has to be considered. The sensing–transmission framework forms a Markov decision process (MDP), and solving the MDP problem exhaustively through conventional methods cannot be a plausible solution for ad hoc networks such as a CRN. In this paper, to solve the MDP problem, an actor–critic-algorithm-based solution for optimizing the action taken in a sensing–transmission framework is proposed. The proposed scheme solves an optimization problem on the basis of the actor–critic algorithm, and the action that brings the highest reward is selected. The optimal policy is determined by updating the optimization problem parameters. The reward is calculated by the critic component through interaction with the environment, and the value function for each state is updated, which then updates the policy function. Simulation results show that the proposed scheme closely follows the exhaustive search scheme and outperforms a myopic scheme in terms of average throughput achieved.

We study jamming attacks in the physical layer of multihop cognitive radio networks (MHCRNs) where energy-constrained relays forward information from the source to the destination. Meanwhile, a jammer can transmit interfering signals on a channel such that all ongoing transmissions on this channel will be corrupted. In this paper, all jammers can attack only one of the predefined channels in each time slot. Moreover, they can randomly switch channels to start jamming another channel at the beginning of every time slot. The switching behavior is assumed to follow a Gaussian distribution. Due to limited battery capacity in the relays, energy harvesting is utilized to solve the energy-constrained problem in the cognitive radio network. Subsequently, relays are able to harvest energy from non-radio frequency (non-RF) signals such as solar, wind, or temperature. In this paper, we determine the throughput/delay ratio as a key metric to evaluate the performance in MHCRNs. Owing to the limited battery capacity in the relays and the jamming problem, the source needs to select proper relays and channels for each data transmission frame to optimize overall network performance in terms of end-to-end delay, throughput, and energy efficiency. Therefore, we provide two novel multihop allocation schemes to maximize achievable end-to-end throughput while minimizing delay in the presence of jammers. Through simulation results, we validate the effectiveness of the proposed schemes under multiple jamming attacks in MHCRNs.

Spectrum sensing is of crucial importance in cognitive radio (CR) networks. In this paper, a reliable spectrum sensing scheme is proposed, which uses K‐nearest neighbor, a machine learning algorithm. In the training phase, each CR user produces a sensing report under varying conditions and, based on a global decision, either transmits or stays silent. In the training phase the local decisions of CR users are combined through a majority voting at the fusion center and a global decision is returned to each CR user. A CR user transmits or stays silent according to the global decision and at each CR user the global decision is compared to the actual primary user activity, which is ascertained through an acknowledgment signal. In the training phase enough information about the surrounding environment, i.e., the activity of PU and the behavior of each CR to that activity, is gathered and sensing classes formed. In the classification phase, each CR user compares its current sensing report to existing sensing classes and distance vectors are calculated. Based on quantitative variables, the posterior probability of each sensing class is calculated and the sensing report is classified into either representing presence or absence of PU. The quantitative variables used for calculating the posterior probability are calculated through K‐nearest neighbor algorithm. These local decisions are then combined at the fusion center using a novel decision combination scheme, which takes into account the reliability of each CR user. The CR users then transmit or stay silent according to the global decision. Simulation results show that our proposed scheme outperforms conventional spectrum sensing schemes, both in fading and in nonfading environments, where performance is evaluated using metrics such as the probability of detection, total probability of error, and the ability to exploit data transmission opportunities.

With the long propagation delay of an acoustic signal in underwater communications systems, relay node selection is one of the key design factors, because it significantly improves end‐to‐end delay, thereby improving overall network performance. To this end, the authors propose orthogonal frequency division multiplexing‐based spectrum‐aware routing (OSAR), a scheme in which spectrum sensing is done by an energy detector, and each sensor node broadcasts its local sensing results to all one‐hop nodes via an extended beacon message. Each sensor node then selects nodes that agree on an idle channel, consequentially forming a set of neighbouring nodes. The selection of a relay node is determined by calculating the transmission delay – the source/relay node selected is the one that has the minimum transmission delay from among all nodes in the neighbouring set. To evaluate OSAR, the authors perform extensive simulations via ns‐MIRACLE for different numbers of channels using a BELLHOP model, and evaluate the average delay for different sensor nodes within the considered network. The results show a substantial decrease in delay as the number of sensor nodes increases in the network. In addition, the authors verify that the packet delivery ratio increases with increases in the number of sensor nodes, and prove better performance in the overhead ratio. The authors' simulation results verify that OSAR outperforms existing solutions.

An optimal multi‐slot channel sensing schedule is proposed in this study that considers an opportunistic spectrum access with the awareness of channel switching delay and imperfect sensing. A practical case is considered where channel availability statistics are usually correlated in time slots and in frequency channels. The switching delays between channels, hardware constraints, and collision with other cognitive users are considered to find an optimal sensing order of the channels that maximises throughput of cognitive user. The optimal sensing order is obtained using the partially observable Markov decision process framework. Throughput of cognitive user, with and without channel sensing errors, is analytically derived and for each case an algorithm is developed. The proposed scheme mitigates the effect of channel sensing errors on the throughput. Performance of the proposed scheme is evaluated through simulations by comparing it with the existing schemes in the literature.

Herein the authors consider throughput maximisation for a secondary user (SU) in a full‐duplex cognitive radio network (FD‐CRN) when the SU has two separate antennas and a self‐interference suppression capability. In the FD‐CRN, the SU can simultaneously sense the spectrum throughout the whole time slot and transmit data. They propose algorithms based on brute‐force search and particle swarm optimisation methods to help the SU achieve optimal detection thresholds for spectrum sensing in two different FD‐CRN scenarios. In the first scenario, the SU individually performs spectrum sensing, whereas in the second scenario the SU's sensing results are improved by means of cooperative spectrum sensing. Theoretical and simulation results herein show that, for certain values of the system parameters in the above two scenarios, the system under consideration provides much higher throughput than previously proposed systems in conditions of high‐transmission power or low signal‐to‐noise ratio of the primary signal.

To reduce the bandwidth requirement for reporting channel, the application of maximum-distance generalised quantiser for cooperative spectrum sensing (CSS) in cognitive radio is considered. The maximum-distance generalised quantiser, which uses Matsushita distance, is turn out to be a minimum mean square error quantiser of log-likelihood ratio (LLR) under absence hypothesis. Therefore the maximum-Matsushita distance quantiser CSS is achieved by utilising the well-known Lloyd–Max quantisation algorithm without the requirement of the prior probabilities of primary signal at each user. The probability density function of the LLR of the sensing information is also formulated for designing the quantiser. Simulation results reveal that the proposed quantiser with only few quantisation bits can provide the same sensing performance of the LLR test using raw data when primary signal has low signal-to-noise ratio.

In this paper, we verify goodness-of-fit testing through the use of the Anderson-Darling (AD) test [1] for spectrum sensing in cognitive radio. In [1], it was shown that spectrum sensing based on the AD test outperforms the energy detection method. However, this positive result can only be obtained when the primary signal is assumed to be static during sensing interval, which is a very rare case in cognitive radio. This assumption reduces the generality of the proposed test in [1]. The verification results of the AD test with some more general and practical primary signals in this paper show that the application of the AD sensing scheme for spectrum sensing in cognitive radio is still a challenge and requires further research.

In this paper, we consider multiple quality of service (QoS) support by using modified largest weighted delay first (M-LWDF) discipline in the adaptive resource allocation of the orthogonal frequency division multiplexing (OFDM) multiple access (OFDMA) systems. We propose a M-LWDF (M Andrews et al., Feb. 2001) based on subchannel assignment, which can satisfy different QoS requirements of services including delay-sensitive and data rate-sensitive traffics in OFDMA systems. The simulation shows that the proposed M-LWDF based subchannel assignment provides controlled QoS provisions while maintaining both the stability of scheduling discipline and the flexibility of the multiple QoS support.

In this paper, we propose a QoS-sensitive admission threshold method for the transmission of the non-real-time data packet in voice/data mixed CDMA systems while guaranteeing the QoS requirements of voice and data calls. The proposed admission scheme fully utilizes the remaining resources after serving the voice users for transmitting non-realtime data packet while meeting the QoS requirements for both services. The system performance of the proposed scheme is measured in terms of the average data throughput, the average delay and the average packet loss probability of data traffic while maintaining the outage probability of voice to be less than the predetermined value. In addition, we investigate the effect of the admission threshold level on the system performances.

In practice, the DS-CDMA system is equipped with a finite number of channel elements (CEs) that performs the baseband spread signal processing for a given channel in the base station. In this situation, the call blocking can be caused not only by the insufficient number of channel elements but also by the limit of available traffic channels. We focus on analyzing the effect of the limited number of CEs on the Erlang capacity of multimedia DS-CDMA systems in the reverse link when the CDMA cells are sectorized with 3 sectors. For the performance analysis, a multi-dimensional Markov chain model is developed. As a result, the more CE results in the larger Erlang capacity. However, the Erlang capacity is saturated after a certain value of CEs where the call blocking is mainly caused by the insufficient channels per sector.