Large public cloud infrastructure can utilise power which is generated by a multiplicity of power plants. The cost of electricity will vary among the power plants and each will emit different amounts of carbon for a given amount of energy generated. This infrastructure services traffic that can come from anywhere on the planet. It is desirable, for latency purposes, to route the traffic to the data centre that is closest in terms of geographical distance, costs the least to power and emits the smallest amount of carbon for a given request. It is not always possible to achieve all of these goals so we model both the networking and computational components of the infrastructure as a graph and propose the Stratus system which utilises Voronoi partitions to determine which data centre requests should be routed to based on the relative priorities of the cloud operator.

The use of machine learning (ML) inference for various applications is growing drastically. ML inference services engage with users directly, requiring fast and accurate responses. Moreover, these services face dynamic workloads of requests, imposing changes in their computing resources. Failing to right-size computing resources results in either latency service level objectives (SLOs) violations or wasted computing resources. Adapting to dynamic workloads considering all the pillars of accuracy, latency, and resource cost is challenging. In response to these challenges, we propose InfAdapter, which proactively selects a set of ML model variants with their resource allocations to meet latency SLO while maximizing an objective function composed of accuracy and cost. InfAdapter decreases SLO violation and costs up to 65% and 33%, respectively, compared to a popular industry autoscaler (Kubernetes Vertical Pod Autoscaler).

Binarization is an extreme quantization technique that is attracting research in the Internet of Things (IoT) field, as it radically reduces the memory footprint of deep neural networks without a correspondingly significant accuracy drop. To support the effective deployment of Binarized Neural Networks (BNNs), we propose CBin-NN, a library of layer operators that allows the building of simple yet flexible convolutional neural networks (CNNs) with binary weights and activations. CBin-NN is platform-independent and is thus portable to virtually any software-programmable device. Experimental analysis on the CIFAR-10 dataset shows that our library, compared to a set of state-of-the-art inference engines, speeds up inference by 3.6 times and reduces the memory required to store model weights and activations by 7.5 times and 28 times, respectively, at the cost of slightly lower accuracy (2.5%). An ablation study stresses the importance of a Quantized Input Quantized Kernel Convolution layer to improve accuracy and reduce latency at the cost of a slight increase in model size.

Cloud federations offer numerous advantages such as improved resilience, improved performance, and the prevention of vendor lock‐in. In order to function efficiently, however, they must be supported by automated solutions to deploy cloud resources securely and efficiently. In particular, a system which allows virtual machine (VM) migration from one cloud to another while recording this for charging and security purposes is particularly useful. This paper presents a lightweight framework called BlockchainBus which offers secure VM migration in cloud federations using blockchain. This framework is a VM migration Blockchain ledger which can be used to record these migrations. BlockchainBus framework is implemented using the HyperLedger solution and deployed on the Microsoft Azure cloud platform. The overhead of this system is then evaluated. The performance of the BlockchainBus framework was compared with the overall VM migration time and was determined as 2.36 seconds. This result is shorter than the overall VM migration time of 5 seconds and indicates that the BlockchainBus gives better performance.

Audio/visual recognition and retrieval applications have recently garnered significant attention within Internet-of-Things (IoT) oriented services, given that video cameras and audio processing chipsets are now ubiquitous even in low-end embedded systems. In the most typical scenario for such services, each device extracts audio/visual features and compacts them into feature descriptors, which comprise media queries. These queries are uploaded to a remote cloud computing service that performs content matching for classification or retrieval applications. Two of the most crucial aspects for such services are: (i) controlling the device energy consumption when using the service, (ii) reducing the billing cost incurred from the cloud infrastructure provider. In this paper we derive analytic conditions for the optimal coupling between the device energy consumption and the incurred cloud infrastructure billing. Our framework encapsulates: the energy consumption to produce and transmit audio/visual queries, the billing rates of the cloud infrastructure, the number of devices concurrently connected to the same cloud server, and the statistics of the query data production volume per device. Our analytic results are validated via a deployment with: (i) the device side comprising compact image descriptors (queries) computed on Beaglebone Linux embedded platforms and transmitted to Amazon Web Services (AWS) Simple Storage Service, (ii) the cloud side carrying out image similarity detection via AWS Elastic Compute Cloud (EC2) spot instances, with the AWS Auto Scaling being used to control the number of instances according to the demand.