Akademik Veri Yönetim Sistemi
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, cilt.16, sa.7, ss.4365-4377, 2020 (SCI-Expanded)
The problem of inefficient and unevenly distributed spectrum usage in industrial wireless networks has led to the emergence of the concept of industrial cognitive radio (CR) networks, which have particularly important applications in automotive industry. Industrial CR networks are planned to function in a wide spectrum range; therefore, they have high energy consumption because of frequency switching while other wireless technologies do not have this problem. A distinctive feature of this switching cost is that it depends on the wideness between the two frequency bands. In this article, we formulate the joint frequency assignment and scheduling problem for multihop industrial CR networks with a single transceiver by considering varying amounts of energy consumption that occurs while CR devices switch to different frequency bands. Our optimization problem, which we formulate as an integer linear program, minimizes the energy cost related to frequency switching while making frequency and time slot allocation to the cognitive devices. We prove that even on star graphs, our formulated problem is inapproximable within any polynomial-time computable function $f(n)$ in addition to being $\textsc {NP}\text {-Hard}$ in the strong sense. Therefore, we propose a polynomial-time heuristic algorithm to solve the energy consumption problem due to channel switching. Simulation results demonstrate that the performance of our heuristic algorithm is very close to the results obtained from the integer linear programming implementation by CPLEX optimization software. We also compare our proposed method with the corresponding constant energy consumption for frequency switching case and two state-of-the-art algorithms and demonstrate that taking into account the different energy consumption while switching to different frequency bands is vital for joint frequency assignment and scheduling in multihop industrial CR networks with a single transceiver.