Title: Designing A Spectrally Efficient and Energy Efficient Data Aided Demand Driven Elastic Architecture for future Networks (SpiderNET).
Funding: This project is funded by the National Science Foundation of USA, under Award No. 1923669.
Duration: Three years (September 1, 2019 - August 31, 2023)
Funding Amount: $ 500,079.00

The problem addressed by the project

The legacy base station centric cellular architecture is marked by tight interlock between the spectral efficiency (SE) and energy efficiency (EE) due to the following reasons:

1. An optimal base station (BS) density that maximizes network energy efficiency is different from the one required to maximize spectral efficiency.
2. For a given BS density, traffic and BS type (and hence BS power consumption model), there exists an optimal set of BS transmission powers that maximizes SE, but a different set of optimal transmission powers is needed to maximize network EE.
3. Increasing BS density increases the network energy consumption. But, in terms of user equipment (UE), on one hand, it decreases energy consumption of UE for uplink (UL) transmission by bringing the BSs closer to the UE on average whereas, on the other hand, it increases the UE battery consumption for cell discovery, signaling and handovers.
   In short, existing cellular architecture’s rigid cell-centric always ON modus operandi simply does not have the degrees of freedoms and adaptability to maximize both SE and EE simultaneously, proactiveness and intelligence to offer the desired level of QoE expected from future networks. Therefore, there is need for a new architecture that can introduce new degrees of freedom to relax the coupling between the between SE, EE to allow joint optimization of both without compromising QoE.

Project Goals

The overarching goal of this project is to design, characterize, optimize and validate a new architecture that enables the additional degrees of freedom in the design and operation of the mobile network to yield substantial gains in both SE and EE while ensuring customizable Quality of Experience (QoE). To this end, we propose a new architecture called SpiderNET: Spectrally Efficient and Energy Efficient Data Aided Demand Driven Elastic Architecture for Future Networks. The key idea behind SpiderNET is to introduce additional degrees of freedom through an intelligent and adaptive operation to relax the rigid SE-EE tradeoff and thus enable simultaneous enhancement of both SE and EE. This is done by shifting the pivot of operation from the rigid always on base station centric cells to user-centric on demand cells. To enable this shift, SpiderNET consists of a layer of low-density large footprint control base station (CBS) underlayed by high-density switchable data base stations (DBS). Both layers of cells are equipped with a database of measurements to train machine learning models and algorithms for enabling proactive, agile and intelligent operation. By switching on/off the DBS the size of user-centric cells (S-Zones) and other parameters are orchestrated proactively by a machine learning based self-organizing Network (SON) engine that leverages the database of selected measurements at DBS and CBS. The size of the S-Zone, density of DBS, transmission powers of DBS, selection between macro diversity and cooperation diversity among DBS in each S-Zone, the contents, the size and spatiotemporal resolution of the database thus provide the new degrees of freedom, intelligence and adaptability. The preliminary studies suggest SpiderNET can substantially enhance both SE and EE without compromising QoE.
This project’s objective is to pave the way for transforming SpiderNET from an idea into functional cellular architecture. To achieve this objective the project has the following three interlinked goals:

1. The first goal in this project is to develop analytical and simulation models to fully characterize the SE and EE of SpiderNET to determine the key design parameters that can be optimized to maximize its SE and EE gains. These models will then be leveraged to design algorithms for maximizing SE and EE in face of changing traffic conditions.
2. The second goal is to design the database of measurements and leveraging this data at DBS and CBS to develop algorithms for proactive cell discovery and selection and radio resource allocation for jointly maximizing both SE and EE without compromising QoE. A key challenge to be addressed in this goal is quantification of the impact of imperfection in database on SpiderNET performance.
3. The third goal is to conduct evaluation and validation of SpiderNET architecture via system level simulations and TurboRAN – an NSF funded end-to-end programmable testbed purpose designed for enabling proposed research.