Machine learning (ML) is expected to solve many challenges in the fifth generation (5G) of mobile networks. However, ML will also open the network to several serious cybersecurity vulnerabilities. Most of the learning in ML happens through data gathered from the environment. Un-scrutinized data will have serious consequences on machines absorbing the data to produce actionable intelligence for the network. Scrutinizing the data, on the other hand, opens privacy challenges. Unfortunately, most of the ML systems are borrowed from other disciplines that provide excellent results in small closed environments. The resulting deployment of such ML systems in 5G can inadvertently open the network to serious security challenges such as unfair use of resources, denial of service, as well as leakage of private and confidential information. Therefore, in this article we dig into the weaknesses of the most prominent ML systems that are currently vigorously researched for deployment in 5G. We further classify and survey solutions for avoiding such pitfalls of ML in 5G systems.
In this paper, a novel method combining cooperative spectrum sensing with quantized soft decision combining is introduced. In order to allow cognitive radios and cognitive networks to opportunistically use spectrum, it is a prerequisite that the license owner or primary user of the spectrum will not be harmfully interfered and the spectrum band will be vacated as soon as the primary user starts its own transmission. There are results indicating that the reliability of sensing information can be improved by exploiting the spatial dimension via cooperation between cognitive radios. Our approach is to further improve the reliability by sharing sensing information between cooperative radios using quantized soft decision combining. Simulations are conducted for the proposed two bit quantized soft decision combining, hard decision combining and nonquantized soft decision combining in an additive white Gaussian noise (AWGN) channel using Welch's periodogram. Hard decision combining is considered with three different decision making rules and the obtained simulation results are verified with analytical performance results for Welch's periodogram. The results show substantial improvement in the detection probability when sensing information between cooperating nodes is shared using two bits instead of one. By using an additional bit it is possible to reach detection probabilities that in hard decision combining would have required one or more additional cooperative users. The results also indicate that the increase in probability of detection is not as significant when full observation of the signal energy is shared between cooperative radios instead of two bits. Thus, almost all the achievable benefit from soft decision combining can be obtained with the proposed quantized soft decision combining.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.