Point-of-Interest (POI) recommendation is significant in location-based social networks to help users discover new locations of interest. Previous studies on such recommendation mainly adopted a centralized learning framework where check-in data were uploaded, trained and predicted centrally in the cloud. However, such a framework suffers from privacy risks caused by check-in data exposure and fails to meet real-time recommendation needs when the data volume is huge and communication is blocked in crowded places. In this paper, we propose PREFER, an edge-accelerated federated learning framework for POI recommendation. It decouples the recommendation into two parts. Firstly, to protect privacy, users train local recommendation models and share multi-dimensional user-independent parameters instead of check-in data. Secondly, to improve recommendation efficiency, we aggregate these distributed parameters on edge servers in proximity to users (such as base stations) instead of remote cloud servers. We implement the PREFER prototype and evaluate its performance using two real-world datasets and two POI recommendation models. Extensive experiments demonstrate that PREFER strengthens privacy protection and improves efficiency with little sacrifice to recommendation quality compared to centralized learning. It achieves the best quality and efficiency and is more compatible with increasingly sophisticated POI recommendation models compared to other state-of-the-art privacy-preserving baselines.

In this paper, we propose DeCNN, a more effective inference approach that uses decoupled CNN structure to optimize model parallelism for distributed inference on end-user devices. DeCNN is novel consisting of three schemes. Scheme-1 is structure-level optimization. It exploits group convolution and channel shuffle to decouple the original CNN structure for model parallelism. Scheme-2 is partition-level optimization. It is based on channel group to partition the convolutional layers, and then leverages input-based method to partition the fully connected layers, further exposing high degree of parallelism. Scheme-3 is communication-level optimization. It uses inter-sample parallelism to hide communications for better performance and robustness, especially in the weak network connections. We use ImageNet to evaluate the effectiveness of DeCNN for image classification on a distributed multi-ARM platform. Notably, when using the number of devices from 1 to 4, DeCNN can accelerate the inference of large-scale ResNet50 by 3.21x, and reduce 65.3% memory footprint, with 1.29% accuracy improvement.

As a decentralized training paradigm, Federated learning (FL) promises data privacy by exchanging model parameters instead of raw local data. However, it is still impeded by the resource limitations of end devices and privacy risks from the ‘curious’ cloud. Yet, existing work predominately ignores that these two issues are non-orthogonal in nature. In this article, we propose a joint design (i.e., AHFL) that accommodates both the efficiency expectation and privacy protection of clients towards high inference accuracy. Based on a cloud-edge-end hierarchical FL framework, we carefully offload the training burden of devices to one proximate edge for enhanced efficiency and apply a two-level differential privacy mechanism for privacy protection. To resolve the conflicts of dynamical resource consumption and privacy risk accumulation, we formulate an optimization problem for choosing configurations under correlated learning parameters (e.g., iterations) and privacy control factors (e.g., noise intensity). An adaptive algorithmic solution is presented based on performance-oriented resource scheduling, budget-aware device selection, and adaptive local noise injection. Extensive evaluations are performed on three different data distribution cases of two real-world datasets, using both a networked prototype and large-scale simulations. Experimental results show that AHFL relieves the end's resource burden (w.r.t. computation time 8.58% $\downarrow$ , communication time 59.35% $\downarrow$ and memory consumption 43.61% $\downarrow$ ) and has better accuracy (6.34% $\uparrow$ ) than 3 typical baselines under the limited resource and privacy budgets. The code for our implementation is available at https://github.com/Guoyeting/AHFL .

Routing algorithms for cache-coherent NoCs only have limited VCs at their disposal, which poses challenges to the design of routing algorithms. Existing fully adaptive routing algorithms apply conservative VC re-allocation: only empty VCs can be re-allocated, which limits performance. We propose two novel flow control designs. First, whole packet forwarding (WPF) re-allocates a nonempty VC if the VC has enough free buffers for an entire packet. WPF does not induce deadlock if the routing algorithm is deadlock-free using conservative VC re-allocation. It is an important extension to several deadlock avoidance theories. Second, we extend Duato's theory to apply aggressive VC re-allocation on escape VCs without deadlock. Finally, we propose a design which maintains maximal routing flexibility with low hardware cost. For synthetic traffic, our design performs averagely 88.9 percent better than existing fully adaptive routing. Our design is superior to partially adaptive and deterministic routing.

Smart sensor-equipped mobile devices sense, collect, and process data generated by the edge network to achieve intelligent control, but such mobile devices usually have limited storage and computing resources. Mobile cloud storage provides a promising solution owing to its rich storage resources, great accessibility, and low cost. But it also brings a risk of information leakage. The encryption of sensitive data is the basic step to resist the risk. However, deploying a high complexity encryption and decryption algorithm on mobile devices will greatly increase the burden of terminal operation and the difficulty to implement the necessary privacy protection algorithm. In this paper, we propose ENSURE (EfficieNt and SecURE), an efficient and secure encrypted search architecture over mobile cloud storage. ENSURE is inspired by edge computing. It allows mobile devices to offload the computation intensive task onto the edge server to achieve a high efficiency. Besides, to protect data security, it reduces the information acquisition of untrusted cloud by hiding the relevance between query keyword and search results from the cloud. Experiments on a real data set show that ENSURE reduces the computation time by 15% to 49% and saves the energy consumption by 38% to 69% per query.

The dot-product $\sum _{i=1}^{N} A_{i}\times B_{i}$ is one of the most frequently used operations for a wide variety of high-performance computing (HPC) and artificial intelligence (AI) applications. However, for large-scale algorithms, such as acrshort GEMM and acrshort FFT, independent additions are necessary to accumulate the results of length-limited dot-product in order to form the final result, thus increasing latency and overhead. Hence, we proposed a dot-product-dual-accumulate (DPDAC) architecture capable of performing $\left({\sum _{i=1}^{N=1,2,4} A_{i}\times B_{i} + \sum _{j=1}^{M=1,2} C_{j}}\right)$ on a wide range of formats. The proposed architecture supports both single-path and dual-path execution. The single path is designed for performing acrshort DP acrshort FMA or DPDAC of lower formats, while dual-path supports parallel operations for single-precision (SP) addition and 2-term SP or acrshort TF32 dot-product or 4-term acrshort HP or BF16 dot-product. Moreover, numerical precision conversion is also supported by the proposed architecture, allowing for the conversion of numbers to higher or lower formats. The proposed DPDAC has been demonstrated to significantly reduce the overhead in comparison to discrete designs that utilize multiple single-mode acrshort FP units to achieve the same functionalities. Furthermore, when compared to the state-of-the-art multiple-precision designs, the proposed architecture has been shown to support a wide range of formats and a greater variety of operations with lower costs.