Browsing by Subject "Spectrum sharing"
Now showing 1 - 3 of 3
- Results Per Page
- Sort Options
Item A deep learning approach to wireless system design for channel sensing, contention & estimation(2022-07-15) Doshi, Akash Sandeep; Andrews, Jeffrey G.; de Veciana, Gustavo; Dimakis, Alexandros; Kim, Hyeji; Yoo, TaesangDeep Learning techniques are expected to play a key role in the development of wireless systems at the Physical (PHY) and Medium Access Control (MAC) layer for sixth generation (6G) communication networks. In particular, learning-based advancements would be needed to provide for (a) more efficient utilization of shared spectrum to accommodate an ever-increasing number of wireless devices and (b) improved scalability of existing signal processing techniques as the spatial and frequency dimensions of wireless architectures rapidly expand. In the first part of this dissertation, we propose a multi-agent deep reinforcement learning (RL) framework to perform contention-based medium access in shared spectrum. Centralized approaches to spectrum sharing require excessive real-time overhead messaging and obtaining the solution is known to be NP-hard. Instead, we assume that base stations operate in a fully decentralized fashion, and model shared spectrum access of base stations performing spectrum sensing as a decentralized partially observable Markov decision process (MDP). We introduce a two-stage MDP in each time slot that uses information from spectrum sensing and reception quality to make a medium access decision. Our distributed reinforcement learning framework achieves performance competitive with a genie-aided adaptive energy detection threshold. We then extend this framework to a larger action space of medium access and transmit modulation scheme. We modify the reward function to provide for this extension and utilize a stabilizing reinforcement learning technique to provide for scalability, hence achieving an improved cumulative reward on both indoor and outdoor layouts with a large number of BSs. In the second part of this dissertation, we develop a deep generative learning framework to perform channel estimation from an insufficient number of pilots in high dimensional wireless systems. Channel estimation using generative priors assumes that the reconstructed channel lies in the range of the generative model, and optimizes the input vector to estimate the channel matrix. We show that our approach outperforms state-of-the-art compressed sensing (CS) baselines. Subsequently, we develop a novel over-the-air design for training the aforementioned deep generative models using Generative Adversarial Networks from pilot measurements instead of clean channel realizations, and still achieve performance competitive with the CS baselines.Item Association and spectrum sharing in cellular networks(2016-12) Gupta, Abhishek K.; Andrews, Jeffrey G.; Heath, Robert W., Jr, 1973-; Baccelli, Francois; Shakkottai, Sanjay; Vikalo, Haris; Visotsky, EugeneMany models have been proposed to evaluate performance of cellular communication systems. However, the emergence of new technologies have changed cellular systems significantly, and requires new modeling and analysis approaches. This dissertation studies network level optimization concerning cell association and spectrum sharing. As the first contribution, the dissertation presents a framework to investigate downlink multi-antenna heterogeneous networks with flexible cell selection and shows that a simple selection bias-based cell selection criterion closely approximates more complex selection rules to maximize mean the signal-to-interference-plus-noise- ratio (SINR). Under this simpler cell selection rule, the exact expressions for coverage probability and achievable rate of a typical user are derived along with an approximation of the coverage optimal cell selection bias. In the second contribution, the dissertation considers a cellular system where users are simultaneously connected to multiple base stations (BSs) to decrease blockage sensitivity and proposes a framework to analyze the correlation in blocking among multiple links. It evaluates the gains of macro-diversity in the presence of random blockages along with the impact of the blockage size. In the third contribution, the dissertation considers spectrum sharing among millimeter wave (mmWave) operators. A two-level architecture is proposed to model a mmWave multi-operator system and the SINR and per-user rate distribution are derived in the presence of spectrum and infrastructure sharing. It is shown that due to narrow beams, license sharing among operators improves system performance by increasing the per-user rate, even when there is no explicit coordination. In the fourth contribution, this analysis is extended to include static coordination among operators in the form of secondary licensing. A framework is developed to model a mmWave cellular system with a primary operator that has an ``exclusive-use'' license with a provision to sell a restricted secondary license to another operator that has a maximum allowable interference threshold. This licensing approach provides a way of differentiating the spectrum access for the different operators. Results show that compared to uncoordinated sharing, a reasonable gain can be achieved using the proposed secondary licensing, especially for edge rates.Item Integrated cellular and device-to-device networks(2014-12) Lin, Xingqin; Andrews, Jeffrey G.Device-to-device (D2D) networking enables direct discovery and communication between cellular subscribers that are in proximity, thus bypassing the base stations (BSs). In principle, exploiting direct communication between nearby mobile devices will improve spectrum utilization, overall throughput, and energy consumption, while enabling new peer-to-peer and location-based applications and services. D2D-enabled broadband communication technology is also required by public safety networks that must function when cellular networks are not available. Integrating D2D into cellular networks, however, poses many challenges and risks to the long-standing cellular architecture, which is centered around the BSs. This dissertation identifies outstanding technical challenges in D2D-enabled cellular networks and addresses them with novel models and fundamental analysis. First, this dissertation develops a baseline hybrid network model consisting of both ad hoc nodes and cellular infrastructure. This model uses Poisson point processes to model the random and unpredictable locations of mobile users. It also captures key features of multicast D2D including multicast receiver heterogeneity and retransmissions while being tractable for analytical purpose. Several important multicast D2D metrics including coverage probability, mean number of covered receivers per multicast session, and multicast throughput are analytically characterized under the proposed model. Second, D2D mode selection which means that a potential D2D pair can switch between direct and cellular modes is incorporated into the hybrid network model. The extended model is applied to study spectrum sharing between cellular and D2D communications. Two spectrum sharing models, overlay and underlay, are investigated under a unified analytical framework. Analytical rate expressions are derived and applied to optimize the design of spectrum sharing. It is found that, from an overall mean-rate perspective, both overlay and underlay bring performance improvements (vs. pure cellular). Third, the single-antenna hybrid network model is extended to multi-antenna transmission to study the interplay between massive MIMO (multi-input multiple-output) and underlaid D2D networking. The spectral efficiency of such multi-antenna hybrid networks is investigated under both perfect and imperfect channel state information (CSI) assumptions. Compared to the case without D2D, there is a loss in cellular spectral efficiency due to D2D underlay. With perfect CSI, the loss can be completely overcome if the number of canceled D2D interfering signals is scaled appropriately. With imperfect CSI, in addition to pilot contamination, a new asymptotic underlay contamination effect arises. Finally, motivated by the fact that transmissions in D2D discovery are usually not or imperfectly synchronized, this dissertation studies the effect of asynchronous multicarrier transmission and proposes a tractable signal-to-interference-plus-noise ratio (SINR) model. The proposed model is used to analytically characterize system-level performance of asynchronous wireless networks. The loss from lack of synchronization is quantified, and several solutions are proposed and compared to mitigate the loss.