End to End Protocol Design for Dynamic Spectrum Access Networks
Abstract
Objective:The past decade has seen extensive research in the lower layers of the protocol stack, resulting in significant enhancements to the science of spectrum sensing for dynamic spectrum access (DSA) networks. While spectrum in the lower MHz range will continue to be extremely valuable for long-range propagation, vast tracts of underutilized bandwidth in new spectrum bands will require completely rethinking the principles of shared medium access and user-centric quality of service (QoS). This proposal describes a holistic research agenda concerned with developing protocol designs, network management algorithms and rigorous analytic frameworks for reliable end-to-end data delivery under assumptions of diverse spectrum bands and different devices with varying levels of access priority and mobility.Approach:The proposaed approach follows a cohesive researchplan, which will significantly impact future tactical communications and networking. The network architecture is composed of DSA enabled access points (APs) that operate via directives from a wireless softwaredefined networking (WSDN) control plane. Each AP can select one of many, or aggregate, spectrum bandsover a range of few MHz up to mmWave range frequencies (60GHz). Furthermore each AP can flexibly allocate a subset of its many antennas for enhanced MIMO operation. The nodes that connect to the APs canbe both stationary devices, or highly mobile units (such as UAVs), with different types of traffic. All these devices are considered as secondary users of the spectrum, or SUs, and must vacate the spectrum upon the return of the licensed/primary users or PUs. The research will focus on ultra-wide spectrum aggregation and adaptive MIMO; streaming and contention-based spectrum access at the link layer; Joint spectrum access and mobility optimization; service-oriented transport layer design; and deep learning-based network management to address these operational challenges.ONR Mission / Relevance:The project will result in new methods and algorithms for managing coexistence will ensure minimized service disruptions to both the aggregating radio as well as conventional wireless equipment in the vicinity. The UAV-centric control and communication framework will enhance the Navy~s capability and scale back the involvement of human operators. As the Navy expands its capabilities to driver-less vehicles, deep-learningbased network management will signal a complete departure from deterministic optimization theory,moving towards self-evolving intelligent radio entities.Short Work Statement:The proposed research will be undertaken through the completion of five concrete goals, as describedbelow.Goal 1: Ultra-wide spectrum aggregation and adaptive MIMO. Solve the spectrum availability problem through a two-fold approach- (i) identify spectrum beyond the assigned or formally licensed frequencies so that discontinuous spectrum chunks can be aggregated, and (ii) increase spatial re-use of the spectrum that has beenchosen through need-based MIMO reconfiguration.Goal 2: Streaming and contention-based spectrum access at the link layer. The research proposes to minimize the downtime for the streaming SUs by judiciously selecting the minimum number ofbackup channels so that efficiency of spectrum utilization is also maximized. For contention-basedrandom access, we aim to understand the optimal packing of multiple SUs in the same licensedchannel, so that the target QoS metrics of per-packet latency and throughput are satisfied.Goal 3: Joint spectrum access and mobility optimizationWhen nodes are highly mobile, maintaining a connected mesh, as well as providing wireless coveragethrough spectrum availability is a challenging task. In this goal, the proposal considers a special case of aswarm of UAVs that can be flexibly deployed to areas of interest. The challenge is to ensure thatthey remain connected to each other, forming an aerial mesh while also maintaining coordinatedspatial coverage.Goa
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 12, 2016
- Source ID
- N000141612651
Entities
People
- Kaushik Roy Chowdhury
Organizations
- Northeastern University
- Office of Naval Research
- United States Navy