Digital Signal Processing


On the In-Band Full-Duplex Gain Scalability in On-demand Spectrum Wireless Local Area Networks

Authors: Saulo Queiroz

The advent Self-Interference Cancellation (SIC) techniques has turned in-band Full-Duplex (FD) radios into a reality. FD radios doubles the theoretical capacity of a half-duplex wireless link by enabling simultaneous transmission and reception in the same channel. A challenging question raised by that advent is whether it is possible scale the FD gain in Wireless Local Area Networks (WLANs). Precisely, the question concerns on how a random access Medium Access Control (MAC) protocol can sustain the FD gain over an increasing number of stations. Also, to ensure bandwidth resources match traffic demands, the MAC protocol design is also expected to enable On-Demand Spectrum Allocation (ODSA) policies in the presence of the FD feature. In this sense, we survey the related literature and find out a coupled FD-ODSA MAC solution lacks. Also, we identify a prevailing practice for the design of FD MAC protocols we refer to as the 1:1 FD MAC guideline. Under this guideline, an FD MAC protocol ‘sees’ the whole FD bandwidth through a single FD PHYsical (PHY) layer. The protocol attempts to occupy the entire available bandwidth with up to two arbitrary simultaneous transmissions. With this, the resulting communication range impair the spatial reuse offer which penalizes network throughput. Also, modulating each data frame across the entire wireless bandwidth demands stronger Received Signal Strength Indication (RSSI) (in comparison to narrower bandwidths). These drawbacks can prevent 1:1 FD MAC protocols to scale the FD gain. To face these drawbacks, we propose the 1:N FD MAC design guideline. Under the 1:N guideline, FD MAC protocols ‘see’ the FD bandwidth through N >1 orthogonal narrow-channel PHY layers. Channel orthogonality increases spatial reuse offer and narrow channels relaxes RSSI requisites. Also, the multi-channel arrangement we adopt facilitates the development of ODSA policies at the MAC layer. To demonstrate how an FD MAC protocol can operate under the 1:N design guideline, we propose two case studies. A case study consists of a novel random access protocol under the 1:N design guideline called the Piece by Piece Enhanced Distributed Channel Access (PbP-EDCA). The other case study consists in adapting an existing FD Wi-Fi MAC protocol [Jain et al., 2011]) – we name as the 1:1 FD Busy Tone MAC protocol (FDBT) – to the 1:N design guideline. Through analytical performance evaluation studies, we verify the 1:N MAC protocols can outperform the 1:1 FDBT MAC protocol’s saturation throughput even in scenarios where 1:1 FDBT is expected to maximize the FD gain. Our results indicate that the capacity upper-bound of an arbitrary 1:1 FD MAC protocol improves if the protocol functioning can be adapted to work under the 1:N MAC design guideline. To check whether that assertion is valid, we propose an analytical study and a proof-of-concept software-defined radio experiment. Our results show the capacity upper-bound gains of both 1:1 and 1:N design guidelines corresponds to 2× and 2.2× the capacity upper-bound achieved by a standard half-duplex WLAN at the MAC layer, respectively. With these results, we believe our proposal can inspire a new generation of MAC protocols that can scale the FD gain in WLANs.

Comments: 126 Pages. PhD Thesis

Download: PDF

Submission history

[v1] 2019-04-24 10:45:49

Unique-IP document downloads: 29 times is a pre-print repository rather than a journal. Articles hosted may not yet have been verified by peer-review and should be treated as preliminary. In particular, anything that appears to include financial or legal advice or proposed medical treatments should be treated with due caution. will not be responsible for any consequences of actions that result from any form of use of any documents on this website.

Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.

comments powered by Disqus