Enhanced Rank Adaptation for MIMO Systems

Multiple-input-multiple-output systems enhance the available degrees of freedom, enable spatial multiplexing, and improve network capacity. However, higher-rank transmission becomes advantageous only if it compensates for inter-layer interference, necessitating rank adaptation. Typically, rank adaptation relies on full or partial channel state information or outer loop rank adaptation (OLRA), with the former incurring additional resource overhead, while the existing works on latter often yield sub-optimal performance. Addressing this issue, we have designed an enhanced OLRA (eOLRA) that achieves optimal performance. We have performed extensive system-level simulations and experimental evaluations on real-time 5G base station to quantify the achievable gains. We demonstrate signficant gains in the cell-edge throughput and link-stability.

More details on this work are available at this link [OLRA]

Graph Neural Networks-based User Pairing in Wireless Communication Systems 

Recently, deep neural networks have emerged as a solution to solve NP-hard wireless resource allocation problems in real-time. However, multi-layer perceptron (MLP) and convolutional neural network (CNN) structures, which are inherited from image processing tasks, are not optimized for wireless network problems. As network size increases, these methods get harder to train and generalize. User pairing is one such essential NP-hard optimization problem in wireless communication systems that entails selecting users to be scheduled together while minimizing interference and maximizing throughput. In this paper, we propose an unsupervised graph neural network (GNN) approach to efficiently solve the user pairing problem. Our proposed method utilizes the Erdos goes neural pipeline to significantly outperform other scheduling methods such as k-means and semi-orthogonal user scheduling (SUS). At 20 dB SNR, our proposed approach achieves a 49% better sum rate than k-means and a staggering 95% better sum rate than SUS while consuming minimal time and resources. The scalability of the proposed method is also explored as our model can handle dynamic changes in network size without experiencing a substantial decrease in performance. Moreover, our model can accomplish this without being explicitly trained for larger or smaller networks facilitating a dynamic functionality that cannot be achieved using CNNs or MLPs 

This work has significant contributions from my colleague Sharan Mourya. More details on this work are available at this link [GNN]. 

Building Bharat RAN 5G

At WiSig Networks, we are developing a trusted, secure, dis-aggregated and standard compliant 5G Bharat RAN solution for Indian market that inter-operates with 5G Core and third-party low-power and high-power radio units. This is currently under outdoor trials with leading public operator in India. This solution can be adapted for global markets including macro as well as private 5G deployments. Some of the key research and development activities of my team includes following: design, implementation, and on-field evaluation of scheduling algorithms, link and rank adaptation, power control, and multi-user MIMO.