Talks and Conference Papers

Minimizing Clearing Time in mmWave Networks with Overlapping Coverage

June 24, 2024

Conference Publication, IEEE Vehicular Technology Conference (VTC) Spring 2024, Singapore.

Abstract: This paper considers millimeter-wave (mmWave) networks with hybrid beamforming communications, where base stations have a limited number of radio frequency (RF) chains. The base stations have overlapping coverage to overcome blockage issues in both downlink and uplink transmission. We propose a user association (UA) scheme that minimizes the time required for clearing data traffic of users in the coverage area. We formulate the UA problem as a time allocation problem, allocating time to user-base station links. We provide an innovative two-stage approach to solve this problem. Stage one optimizes a time fraction allocation for user-base station links. Then these time fractions are distributed across the RF chains at each base station using a fully distributed algorithm. Stage two then schedules the user-base station links, provably solving the UA minimum clearing time problem. We then characterize the achievability of any set of target user rates. Numerical results show that our proposed UA scheme achieves significantly reduced clearing times in comparison to baseline schemes.

Short Message Success Rate for LEO Satellite IoT Data Harvesting

June 09, 2024

Conference Publication, IEEE International Conference on Communications 2024, Denver, Colorado, USA.

Abstract: This paper analyses the data message success rate for Internet of Things (IoT) sensing devices communicating over Low Earth Orbit (LEO) satellite links. We present an analytical framework for optimizing multi-objective multi-packet reception on the uplink. We present an analytical result for the probability of message success for a given ground terminal, and present an analytical result for the overall probability of message success, averaged across all terminals.

Zak-OTFS Implementation using Time and Frequency Windowing.

February 21, 2024

Technical Talk, Information Theory and Applications Workshop (ITA) 2024, San Diego, California, USA.

Abstract: Zak-OTFS is a new delay-Doppler modulation technique which uses Zak domain basis functions for signaling, and requires a new filtering technique for pulse shaping called delay-Doppler domain twisted convolution (TC) filtering. In this talk, we show how to implement these filters in the time domain and generate the corresponding time domain waveforms. We introduce two general classes of TC filters (Type-1 and Type-2), and show that they can be realized by time and frequency windowing. We then propose practical methods to generate time domain signals for these TC filters. For Type-1, the signals are generated using an interpolation filter. For Type-2, they are generated using a form of precoded OFDM. We show that the Type-2 signals are more spectrally efficient than their Type-1 counterparts.

Slot reservation framework for joint flow control and resource allocation in mmWave IAB networks

June 21, 2023

Technical talk, 2023 Australian Communication Theory Workshop, Darwin, NT, Australia.

Abstract: The challenge of cross-layer control in wireless networks is to optimally control flow rates in conjunction with scheduling links. This is an NP Hard problem in general due to the hardness of the scheduling component. It was recently shown that maximum weighted schedule can be computed efficiently in mmWave integrated access and backhaul (IAB) networks due to the tree structure. Hence in our recent work, we have considered the problem of joint resource allocation and flow control for mmWave IAB networks. Our new approach to scheduling is based around solving a minimum clearing time linear program. We develop the Static Slot Reservation (SSR) algorithm which can be described as a book ahead slot reservation system that is completely distributed and solves the minimum clearing time linear program. We develop our approach further in the Dynamic Slot Reservation (DSR) algorithm, by jointly controlling flow rates and slot reservation in a dynamic IAB network with stochastic UE arrivals and departures. The proposed DSR framework leads to uniform end-to-end delays for flows and can also be implemented to provide end-to-end QoS guarantees. In this talk, we provide an overview of our new distributed slot reservation framework.

This talk is based on joint work with Prof. Philip Whiting and Prof. Stephen Hanly.

Greedy Scheme for Optimal Resource Allocation in HetNets with Wireless Backhaul

June 04, 2017

Conference Publication, IEEE Vehicular Technology Conference (VTC) Spring 2017, Sydney, Australia.

Abstract: We formulate a linear programming problem to find the minimum clearing time in HetNets. Although this program is NP hard in general, we consider particular topologies that arise in HetNets, including a two cell HetNet and a linear chain of HetNets, both with wireless backhaul, and we provide an efficient, greedy algorithm that provably solves the minimum clearing time problem for these networks. We show how this algorithm can be applied to jointly optimize the ABS time across multiple macros in a HetNet, and we demonstrate capacity gains of the algorithm, compared to standard approaches to Inter Cell Interference Coordination. This paper provides insight into how to manage interference in presence of more than one macro, and how to efficiently operate wireless backhaul in HetNets.