Sensors mounted in infrastruture  such as LIDARs, RADARs and Cameras can provide huge amount of data as situation awareness information that can be utilized to enhance wireless network quality.   Accurate user positioning (cm range), Vision aided beamforming, Blockage prediction and proactive handovers for mmWave communications, received power prediction for both outdoor and indoor scenarios are major research directions.

The learning capability of deep learning-based methods are exploited to perform the resource allocation tasks in cell-free massive MIMO networks in place of conventional optimization-based approaches. Deep neural networks are trained to find optimal power allocations, fronthaul capacity allocations etc. to achieve a given objective such as maximizing the minimum user rate or sum rate of the network.

Deep neural network-based active user detection (AUD) schemes for the grant-free sparse code multiple access (SCMA) system in mMTC uplink framework are proposed. The deep neural network learns the nonlinear mapping between the received signal and the desired support. The offline pre-trained model can detect the active devices without any channel state information and prior knowledge of the device sparsity level. 

The use of an intelligent reflecting surface (IRS) in a mmWave vehicular network is investigated.  An IRS consists of passive elements, which can reflect the incoming signals with adjustable phase shifts. Phase shifts of the elements can be controlled in an intelligent way to improve the performance of communication. We consider the problem of rate maximization in the uplink while utilizing the IRS. Numerical results show the ability of an IRS to significantly improve the performance, which is further validated by simulations performed using a commercial ray-tracing tool.

Polar codes are a class of error-correcting codes that can achieve channel capacity in long block lengths. Fast Simplified Successive Cancellation (FSSC) is the algorithm that makes polar codes practical for the cutting-edge communication systems. Concatenated with a Cyclic Redundancy Check (CRC), list decoders can improve error correction performance.

The main drawback is that their decoding algorithm is serial in nature. Taking into account new developments in machine learning, we are trying to develop efficient algorithms to increase the throughput to address requirements in 6G generation 3GPP standard for New Radio. We finally implement our algorithms on embedded platform and test our systems with hardware platform in the loop to assure full functionality of algorithms. 

High Millimeter wave (mmWave) based multiple-input multiple-output (MIMO) capable user-centric (UC) ultra-dense (UD) networks are suggested to facilitate high throughput requirements of future networks. Due to the high blockage susceptibility of mmWave, the connections may drop frequently. Hence efficient and fast beam management in initial access (IA) is essential. Current cellular systems use beam sweeping based IA mechanisms. UC UD concept requires all of its access points (APs) to perform IA. This leads to a shortage of orthogonal radio resources. Nonorthogonal resource allocation causes interference which leads to a higher misdetection probability. In this paper, we propose a novel deep contextual bandit (DCB) based approach to perform fast and efficient IA in mmWave based UC UD networks. The DCB model uses one reference signal from the user to predict the IA beam. The reduced use of reference signals improves beam discovery delay and relaxes the requirement for radio resources. Ray-tracing and stochastic channel model- based simulations show that the suggested system outperforms its beam sweeping counterpart in terms of probability of beam misdetection and beam discovery delay in mmWave based UC UD networks.