A Time-Varying Opportunistic Multiple Access for Delay-Sensitive Inference in Wireless Sensor Networks

Cohen, Kobi; Malachi, Dean

doi:10.1109/access.2019.2955741

“…Model-free learning strategies were developed in [33], [45], [46] for orthogonal channels, and multiple access channel strategies were developed in [47]- [49]. Graph coloring formulations have dealt with modeling the spectrum access problem as a graph coloring problem, in which users and channels are represented by vertices and colors, respectively (see [38]- [41] and references therein for related studies).…”

Section: B Related Workmentioning

confidence: 99%

Distributed Learning Over Markovian Fading Channels for Stable Spectrum Access

Gafni

¹

,

Cohen

²

2022

Self Cite

View full text Add to dashboard Cite

We consider the problem of multi-user spectrum access in wireless networks. The bandwidth is divided into K orthogonal channels, and M users aim to access the spectrum. Each user chooses a single channel for transmission at each time slot. The state of each channel is modeled by a restless unknown Markovian process. Previous studies have analyzed a special case of this setting, in which each channel yields the same expected rate for all users. By contrast, we consider a more general and practical model, where each channel yields a different expected rate for each user. This model adds a significant challenge of how to efficiently learn a channel allocation in a distributed manner to yield a global system-wide objective. We adopt the stable matching rate as the system objective, which is known to yield strong performance in multichannel wireless networks, and develop a novel Distributed Stable Strategy Learning (DSSL) algorithm to achieve the objective. We prove theoretically that DSSL converges to the stable matching allocation, and the regret, defined as the loss in total rate with respect to the stable matching solution, has a logarithmic order with time. Simulation results demonstrate the strong performance of the DSSL algorithm. We show that DSSL has significant advantages over the state-of-the-art methods in terms of implementation complexity, convergence speed, and network churn conditions.INDEX TERMS Wireless networks, spectrum access, online learning.

show abstract

“…The game theoretic aspects of the problem have been investigated from both non-cooperative (i.e., each user aims at maximizing an individual utility) [18], [19], [24], [25], [33], and cooperative (i.e., each user aims at maximizing a system-wide global utility) [17], [26], [34], [35] settings. Model-free learning strategies were developed in [36], [37] for orthogonal channels, compact models [38], and multiple access channel strategies were developed in [39], [40]. Graph coloring formulations have dealt with modeling the spectrum access problem as a graph coloring problem, in which users and channels are represented by vertices and colors, respectively (see [29]- [32] and references therein for related studies).…”

Section: B Related Workmentioning

confidence: 99%

“…With (50) we can bound (40), and therefore the regret due to sub-optimal allocation in the exploitation phases is bounded by:…”

mentioning

confidence: 99%

Distributed Learning over Markovian Fading Channels for Stable Spectrum Access

Gafni¹,

Cohen²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

We consider the problem of multi-user spectrum access in wireless networks. The bandwidth is divided into K orthogonal channels, and M users aim to access the spectrum. Each user chooses a single channel for transmission at each time slot. The state of each channel is modeled by a restless unknown Markovian process. Previous studies have analyzed a special case of this setting, in which each channel yields the same expected rate for all users. By contrast, we consider a more general and practical model, where each channel yields a different expected rate for each user. This model adds a significant challenge of how to efficiently learn a channel allocation in a distributed manner to yield a global system-wide objective. We adopt the stable matching utility as the system objective, which is known to yield strong performance in multichannel wireless networks, and develop a novel Distributed Stable Strategy Learning (DSSL) algorithm to achieve the objective. We prove theoretically that DSSL converges to the stable matching allocation, and the regret, defined as the loss in total rate with respect to the stable matching solution, has a logarithmic order with time. Finally, simulation results demonstrate the strong performance of the DSSL algorithm. I. INTRODUCTIONWe consider the spectrum access problem, where a shared bandwidth is divided into K orthogonal channels (i.e., subbands), and M users want to access the spectrum, where K ≥ M . Each channel is modeled by a Finite-State Markovian Channel (FSMC), which is independent and non-identically distributed across channels. The FSMC is a tractable model widely used to capture the time-varying behavior of a radio communication channel [2], [3]. It is often employed to model radio channel dynamics due to primary user occupancy effects in hierarchical cognitive radio networks (where the M secondary (unlicensed) users are cognitive in terms of learning and adapting good access strategies), or the external interference effects in the open sharing model among M users in the wireless network (e.g., ISM band) [4], [5]. At each time step, each user experiences a different transmission rate over each channel depending on its FSMC distribution, where the FSMC parameters (i.e., the transition probabilities that govern the Markov chain) are unknown. At each time step, each user is allowed to choose one channel to access, and observe the instantaneous channel state. If two users or more access the same channel at the same time, a collision occurs and the achievable rate is zero.We adopt the stable matching utility (see Section II for details) as the system objective, which is known to yield strong Tomer Gafni and Kobi Cohen are with the School of Electrical and Computer Engineering,

show abstract

“…This strategy builds upon the fact that when the participating users operate over the same wireless network, uplink transmissions are carried out over a mulitple access channel (MAC). Model-dependent inference over MACs is relatively well-studied in the sensor network literature, where methods for model-dependent inference over MACs and theoretical performance guarantees have been established under a wide class of problem settings [23]- [32]. These studies focused on model-based inference, and not on machine learning paradigms.…”

mentioning

confidence: 99%

Over-the-Air Federated Learning from Heterogeneous Data

Sery,

Shlezinger,

Cohen

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Federated learning (FL) is a framework for distributed learning of centralized models. In FL, a set of edge devices train a model using their local data, while repeatedly exchanging their trained updates with a central server. This procedure allows tuning a centralized model in a distributed fashion without having the users share their possibly private data. In this paper, we focus on over-the-air (OTA) FL, which has been suggested recently to reduce the communication overhead of FL due to the repeated transmissions of the model updates by a large number of users over the wireless channel. In OTA FL, all users simultaneously transmit their updates as analog signals over a multiple access channel, and the server receives a superposition of the analog transmitted signals. However, this approach results in the channel noise directly affecting the optimization procedure, which may degrade the accuracy of the trained model. We develop a Convergent OTA FL (COTAF) algorithm which enhances the common local stochastic gradient descent (SGD) FL algorithm, introducing precoding at the users and scaling at the server, which gradually mitigates the effect of the noise. We analyze the convergence of COTAF to the loss minimizing model and quantify the effect of a statistically heterogeneous setup, i.e. when the training data of each user obeys a different distribution. Our analysis reveals the ability of COTAF to achieve a convergence rate similar to that achievable over error-free channels. Our simulations demonstrate the improved convergence of COTAF over vanilla OTA local SGD for training using non-synthetic datasets. Furthermore, we numerically show that the precoding induced by COTAF notably improves the convergence rate and the accuracy of models trained via OTA FL. I. INTRODUCTIONRecent years have witnessed unprecedented success of machine learning methods in a broad range of applications [2]. These systems utilize highly parameterized models, such as deep neural networks (DNNs), trained using massive data sets. In many applications, samples are available at remote users, e.g. smartphones, and the common strategy is to gather these samples at a computationally powerful server, where the model is trained [3]. Often, data sets contain private information, and thus the user may not be willing to share them with the server. Furthermore, sharing massive data sets can result in a substantial burden on the communication links between the users and the server. To allow centralized training without data sharing, federated learning (FL) was proposed in [4] as a method combining A short version of this paper that introduces the algorithm for i.i.d. data and preliminary simulation results was accepted for presentation in the 2020 IEEE Global Communications Conference (GLOBECOM) [1].

show abstract

A Time-Varying Opportunistic Multiple Access for Delay-Sensitive Inference in Wireless Sensor Networks

Cited by 8 publications

References 46 publications

Distributed Learning Over Markovian Fading Channels for Stable Spectrum Access

Distributed Learning Over Markovian Fading Channels for Stable Spectrum Access

Distributed Learning over Markovian Fading Channels for Stable Spectrum Access

Over-the-Air Federated Learning from Heterogeneous Data

Contact Info

Product

Resources

About