Accelerating Beam Sweeping in mmWave Standalone 5G New Radios Using Recurrent Neural Networks

Abstract

Millimeter wave (mmWave) is a key technology to support high data rate demands for 5G applications. Highly directional transmissions are crucial at these frequencies to compensate for high isotropic pathloss. This reliance on directional beamforming, however, makes the cell discovery (cell search) challenging since both base station (gNB) and user equipment (UE) jointly perform a search over angular space to locate potential beams to initiate communication. In the cell discovery phase, sequential beam sweeping is performed through the angular coverage region in order to transmit synchronization signals. The sweeping pattern can either be a linear rotation or a hopping pattern that makes use of additional information. This paper proposes beam sweeping pattern prediction, based on the dynamic distribution of user traffic, using a form of recurrent neural networks (RNNs) called Gated Recurrent Unit (GRU). The spatial distribution of users is inferred from data in call detail records (CDRs) of the cellular network. Results show that the users spatial distribution and their approximate location (direction) can be accurately predicted based on CDRs data using GRU, which is then used to calculate the sweeping pattern in the angular domain during cell search.

Deep Learning for Fast and Reliable Initial Access in AI-Driven 6G mmWave Networks

“…The average discovery time by leveraging knowledge obtained from real time arrival statistics of incoming users was reduced in [18]. Gated recurrent neural networks (RNNs) were used in [19] to predict the sequence of the beams swept during IA. Using call records to find the user's location, [19] used this information to determine the number of users in each sector such that the beams are swept in a descending order of users in each sector, i.e., sector with the highest number of users goes first and so on.…”

“…Gated recurrent neural networks (RNNs) were used in [19] to predict the sequence of the beams swept during IA. Using call records to find the user's location, [19] used this information to determine the number of users in each sector such that the beams are swept in a descending order of users in each sector, i.e., sector with the highest number of users goes first and so on. Pseudo-directional patterns were considered in [19] since the operations are in the sub-6 GHz band.…”

A Comprehensive Survey on Mobility Management in 5G Heterogeneous Networks: Architectures, Challenges and Solutions

“…• In the design of 5G mmWave system channels, it is very important to use real measurement data to make more accurate predictions and provide better performance. [45], [46], [47], [48], [49], [50], [51], [52], [53] Load balancing…”

“…However, sequential search causes large amounts of access latency and low initial access efficiency. Therefore, using the repetitive neural networks (RNN) called gated recurrent unit in [49], a beam sweeping model based on the dynamic distribution of user traffic is presented. Spatial distributions of users are provided from the data in the cellular network call detail records (CDR).…”

Beam Tracking Strategies for 5G New Radio Networks Operating in the Millimetre Wave Bands

“…Authors in [75] use a type of Recurrent Neural Network (NN) (RNN) to predict the beam sweeping pattern according to the users spatial distribution. Their goal is to cover all users with the minimum number of beams, thus reducing the beam sweeping procedure timing.…”