In this work, a new prototype of the eight-element MIMO antenna system for 5G communications, internet of things, and networks has been proposed. This system is based on an H-shaped monopole antenna system that offers 200 MHz bandwidth ranges between 3.4–3.6 GHz, and the isolation between any two elements is well below −12 dB without using any decoupling structure. The proposed system is designed on a commercially available 0.8 mm-thick FR4 substrate. One side of the chassis is used to place the radiating elements, while the copper from the other side is being removed to avoid short-circuiting with other components and devices. This also enables space for other systems, sub-systems, and components. A prototype is fabricated and excellent agreement is observed between the experimental and the computed results. It was found that ECC is 0.2 for any two radiating elements which is consistent with the desirable standards, and channel capacity is 38 bps/Hz which is 2.9 times higher than 4 × 4 MIMO configuration. In addition, single hand mode and dual hand mode analysis are conducted to understand the operation of the system under such operations and to identify losses and/or changes in the key performance parameters. Based on the results, the proposed antenna system will find its applications in modern 5G handheld devices and internet of things with healthcare and high rate delivery. Besides that, its design simplicity will make it applicable for mass production to be used in industrial demands.

Reinforcement learning (RL) is an unsupervised learning technique used in many real-time applications. The essence of RL is a decision-making problem. In RL, the agent constantly interacts with the environment and selects the next action according to previous feedback in terms of reward. In this paper, RL trains Software-Defined Wireless Sensor Networks (SDWSNs) controller to optimize the routing paths. We combine RL and SDN, where RL is applied to the SDN controller to generate the routing tables. We also propose four different reward functions for optimization of the network performance. RL-based SDWSN improves network performance by 23% to 30% in terms of lifetime compared with RL-based routing techniques. RL-based SDWSN performs well because it can intelligently learn the routing path at the controller. In addition, it has a faster network convergence rate than RL-based WSN.

This work presents a design technique to implement miniaturized cross-coupled bandpass filters in inline physical configurations based on triple-mode resonators. Triple-mode resonances are obtained by using integrated coaxial-waveguide cavity resonators. They consist of two coaxial conducting posts placed in the sidewalls of a rectangular waveguide cavity. In the proposed triplet, a transmission zero (TZ) can be positioned at any of the two sides of the passband by simply locating the coaxial posts either on the same wall or on opposite walls of the rectangular waveguide cavity. Coaxial-to-waveguide mode coupling is implemented by means of coupling screws, which adds flexibility in terms of tuning easiness. Three 9.9-GHz proof-of-concept prototypes corresponding to third- and sixth-order integrated coaxial-waveguide bandpass filters are designed at the electromagnetic-simulation level. Furthermore, the triplet with a TZ below the passband is manufactured and characterized to experimentally validate the engineered 3-D RF filter principle and its underlying design theory.

This work presents the design techniques for generalized Chebyshev miniaturized monolithic ceramic bandpass filters. The integrated ceramic dielectric waveguide filters occupy less than 50% volume of conventional coaxial TEM filters. Generalized Chebyshev filter response is achieved using metal-plated blind holes embedded at various places in the dielectric slab, which provide inter-resonating couplings and/or attenuation poles at the same time. The measured results of an equivalent air-filled model are presented for the verification of the concept. The measurements closely agree with the design theory.

Hyperbolic metamaterials (HMMs) exhibit unusual electromagnetic properties, attracting significant interest from both fundamental physics and applied engineering perspectives. At subwavelength dimensions, the effective medium theory (EMT) is commonly used to simulate and describe HMMs, which are typically composed of metal–dielectric multilayers, due to their simplicity in design. In this context, the EMT approximation must align with exact electromagnetic solutions. However, significant discrepancies in the predicted properties sometimes arise due to these approximate methods. This article compares the EMT with the exact solution (i.e., the transfer matrix method) and demonstrates that the transmission and reflection in the HMM multilayer structure are sensitive to variations in the number of layers and the filling fraction, even when the layer thicknesses are much smaller than the operating wavelength. Understanding this dependence of EMT on the number of layers and the filling fraction is crucial for reducing errors and improving the design of plasmonics and photonics devices.