Over-The-Air Computation in Correlated Channels

  title={Over-The-Air Computation in Correlated Channels},
  author={Matthias Frey and Igor Bjelakovic and Sławomir Stańczak},
  journal={2020 IEEE Information Theory Workshop (ITW)},
This paper addresses the problem of Over-The-Air (OTA) computation in wireless networks which has the potential to realize huge efficiency gains for instance in training of distributed ML models. We provide non-asymptotic, theoretical guarantees for OTA computation in fast-fading wireless channels where the fading and noise may be correlated. The distributions of fading and noise are not restricted to Gaussian distributions, but instead are assumed to follow a distribution in the more general… 

Figures from this paper

Private Federated Learning With Misaligned Power Allocation via Over-the-Air Computation

An optimization problem is formulated to minimize the optimality gap of MPA-DPFL while guaranteeing a certain degree of privacy protection and it is demonstrated that the MPA -DPFL is more suitable than the DPFL with aligned power allocation when the channel condition of a device in the system is lower than a threshold.

A Semantics-Aware Communication Scheme to Estimate the Empirical Measure of A Quantity of Interest Via Multiple Access Fading Channels

This paper considers a setting where the goal is to devise a joint sampling and communication scheme over a wireless multiple access channel that enables to compute the empirical probability measure of a quantity of interest (QoI) at the destination.

Distributed Computation over MAC via Kolmogorov-Arnold Representation

Kolmogorov’s representation theorem provides a framework for decomposing any arbitrary realvalued, multivariate, and continuous function into a two-layer nested superposition of a finite number of functions, which provides modularity and universality to the design of the inner function, and subsequently, a part of computation.

Blind Federated Edge Learning

An analog ‘over-the-air’ aggregation scheme, in which the devices transmit their local updates in an uncoded fashion, and the proposed algorithm becomes deterministic despite the lack of perfect CSI when the PS has a sufficiently large number of antennas.

Towards Secure Over-The-Air Computation

A friendly jammer whose signal is -- contrary to common intuition -- stronger at the legitimate receiver than it is at the eavesdropper is used to protect Over-The-Air computation schemes against passive eavesdropping.

Multi-Slot Over-The-Air Computation in Fading Channels

A multi-slot over-the-air computation (MS-AirComp) method for the sum estimation in fading channels that effectively reduces the computation error meanwhile controlling transmission power, compared with state-of- the-art methods.



Over-The-Air Computation for Distributed Machine Learning

This paper considers noisy fast-fading channels that pose major challenges to the ``over-the-air'' computation of a function of data captured at different wireless devices and approximated function values are approximated from superimposed noisy signals transmitted by different devices.

Computation Over Multiple-Access Channels

It is shown that there is no source-channel separation theorem even when the individual sources are independent, and joint source- channel strategies are developed that are optimal when the structure of the channel probability transition matrix and the function are appropriately matched.

Distributed Approximation of Functions over Fast Fading Channels with Applications to Distributed Learning and the Max-Consensus Problem

This work takes fast fading into account and gives explicit probability bounds for the approximation error allowing us to derive bounds on the number of channel uses that are needed to approximate a function up to a given approximation accuracy.

Over-the-Air Computation Systems: Optimization, Analysis and Scaling Laws

This paper considers an optimization problem to minimize the computation mean-squared error of the K sensors’ signals at the receiver by optimizing the transmitting-receiving (Tx-Rx) policy, under the peak power constraint of each sensor.

On Analog Gradient Descent Learning Over Multiple Access Fading Channels

This work develops a novel Gradient-Based Multiple Access (GBMA) algorithm that can approach the convergence rate of the centralized gradient descent (GD) algorithm in large networks, and establishes a finite-sample bound of the error for both convex and strongly convex loss functions with Lipschitz gradient.

Nomographic Functions: Efficient Computation in Clustered Gaussian Sensor Networks

It is shown that at the cost of a reduced rate, a computation scheme can be extended to compute every continuous function of the sensor readings in a finite succession of steps, where in each step a different nomographic function is computed, demonstrating the fundamental role of nomographic representations.

Compute-and-Forward: Harnessing Interference Through Structured Codes

This paper proposes a new strategy, compute-and-forward, that exploits interference to obtain significantly higher rates between users in a network and is applicable even if the transmitters lack channel state information.

Robust Analog Function Computation via Wireless Multiple-Access Channels

This paper proposes an analog computation scheme that allows for an efficient estimate of linear and nonlinear functions over the wireless multiple-access channel and analyses the estimation error for two function examples to show the potential for huge performance gains over time- and code-division multiple- access based computation schemes.

Collaborative Machine Learning at the Wireless Edge with Blind Transmitters

Theoretical analysis indicates that, with the proposed DSGD scheme, increasing the number of PS antennas mitigates the fading effect, and, in the limit, the effects of fading and noise disappear, and the PS receives aligned signals used to update the model parameter.

Blind Over-the-Air Computation and Data Fusion via Provable Wirtinger Flow

This work establishes the global convergence guarantee of Wirtinger flow with random initialization for BlairComp problem, which enjoys a model-agnostic and natural initialization implementation for practitioners with theoretical guarantees.