Broadband Analog Aggregation for Low-Latency Federated Edge Learning

By Guangxu Zhu, Yong Wang, Kaibin Huang. 2018

784 citations recorded by the source corpus.

To leverage rich data distributed at the network edge, a new machine-learning paradigm, called edge learning, has emerged where learning algorithms are deployed at the edge for providing intelligent services to mobile users. While computing speeds are advancing rapidly, the communication latency is becoming the bottleneck of fast edge learning. To address this issue, this work is focused on designing a low-latency multi-access scheme for edge learning. To this end, we consider a popular privacy-preserving framework, federated edge learning (FEEL), where a global AI-model at an edge-server is updated by aggregating (averaging) local models trained at edge devices. It is proposed that the updates simultaneously transmitted by devices over broadband channels should be analog aggregated “over-the-air” by exploiting the waveform-superposition property of a multi-access channel. Such broadband analog aggregation (BAA) results in dramatical communication-latency reduction compared with the conventional orthogonal access (i.e., OFDMA). In this work, the effects of BAA on learning performance are quantified targeting a single-cell random network. First, we derive two tradeoffs between communication-and-learning metrics, which are useful for network planning and optimization. The power control (“truncated channel inversion”) required for BAA results in a tradeoff between the update-reliability [as measured by the receive signal-to-noise ratio (SNR)] and the expected update-truncation ratio. Consider the scheduling of cell-interior devices to constrain path loss. This gives rise to the other tradeoff between the receive SNR and fraction of data exploited in learning. Next, the latency-reduction ratio of the proposed BAA with respect to the traditional OFDMA scheme is proved to scale almost linearly with the device population. Experiments based on a neural network and a real dataset are conducted for corroborating the theoretical results.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Electrical and Electronic Engineering

Citations in the corpus

Age-Based Scheduling Policy for Federated Learning in Mobile Edge Networks (2019, background): “…studies have been carried out and resulted in various scheduling protocols for FL, ranging from minimizing the transmission latency [13], maximizing the spectral utility [14], to opportunistically alternating between selecting the UEs with advantageous and disadvantageous channel conditions [15].”
Scheduling Policies for Federated Learning in Wireless Networks (2019, methodology): “When spectral resources become the communication bottleneck, there are new methods exploiting the computeover-air mechanism and a jointly decode-and-average scheme at the edge computing unit [17], [18].”
Edge Artificial Intelligence for 6G: Vision, Enabling Technologies, and Applications (2021, methodology): “To tackle the channel fading perturbation, a channel inversion method was proposed in [45], [173], [44] by multiplying the inverse of channel gain for the transmit signal, which may however not satisfy the power constraint at edge devices.”
Energy Efficient Federated Learning Over Wireless Communication Networks (2019, methodology): “To minimize latency, a broadband analog aggregation multi-access scheme was designed in [28] for FL by exploiting the waveform-superposition property of a multi-access channel.”
One-Bit Over-the-Air Aggregation for Communication-Efficient Federated Edge Learning: Design and Convergence Analysis (2020, methodology): “A performance comparison between digital transmission using OFDMA, BAA developed in [6], and the proposed OBDA is presented in Fig.”
Distributed Learning in Wireless Networks: Recent Progress and Future Challenges (2021, background): “The main drawback of the classic orthogonalaccess schemes (e.g., OFDMA or TDMA) is that they do not scale well with the number of devices.”
Joint Device Scheduling and Resource Allocation for Latency Constrained Wireless Federated Learning (2020, methodology): “We accept the common assumption that each device has equal amount of training data samples and the local training datasets are non-overlapping with each other [15], [20].”
Edge Learning for B5G Networks With Distributed Signal Processing: Semantic Communication, Edge Computing, and Wireless Sensing (2022, positive): “Besides, a multi-access network structure for EL was proposed in [15] to reduce the communication latency.”
On Analog Gradient Descent Learning Over Multiple Access Fading Channels (2019, background): “In [13], [14], the authors considered transmissions over fading MAC, where each entry of the gradient vector is scheduled for transmission depending on the corresponding channel condition.”
Federated Learning for Vehicular Internet of Things: Recent Advances and Open Issues (2020, background): “In [59], Zhu et al. present a broadband analog aggregation framework for low latency FL in wireless networks.”
Federated Learning via Over-the-Air Computation (2018, background): “This tradeoff between learning and aggregation is also considered in the recent parallel work [30], which quantifies the device population by excluding the devices with weak channel coefficients under deep channel fading and assuming IID located devices.”
Federated Learning Over Wireless Fading Channels (2019, methodology): “In [26] the authors consider a fading MAC, and also apply analog transmission, where each entry of a gradient vector at each of the devices is scheduled for transmission depending on the corresponding channel condition.”
A Comprehensive Overview on 5G-and-Beyond Networks With UAVs: From Communications to Sensing and Intelligence (2020, background): “AirComp leads to a fully-integrated-communication-computation design, and thus can have abundant applications in distributed sensing, computation, federated learning [336], [337], and UAV swarms.”
Convergence of Edge Computing and Deep Learning: A Comprehensive Survey (2019, background): “Aiming to accelerate the global aggregation in FL, [214] takes advantage of over-the-air computation [215]–[217], of”