On lightweight privacy-preserving collaborative learning for internet-of-things objects

Abstract

The Internet of Things (IoT) will be a main data generation infrastructure for achieving better system intelligence. This paper considers the design and implementation of a practical privacypreserving collaborative learning scheme, in which a curious learning coordinator trains a better machine learning model based on the data samples contributed by a number of IoT objects, while the confidentiality of the raw forms of the training data is protected against the coordinator. Existing distributed machine learning and data encryption approaches incur significant computation and communication overhead, rendering them ill-suited for resourceconstrained IoT objects. We study an approach that applies independent Gaussian random projection at each IoT object to obfuscate data and trains a deep neural network at the coordinator based on the projected data from the IoT objects. This approach introduces light computation overhead to the IoT objects and moves most workload to the coordinator that can have sufficient computing resources. Although the independent projections performed by the IoT objects address the potential collusion between the curious coordinator and some compromised IoT objects, they significantly increase the complexity of the projected data. In this paper, we leverage the superior learning capability of deep learning in capturing sophisticated patterns to maintain good learning performance. Extensive comparative evaluation shows that this approach outperforms other lightweight approaches that apply additive noisification for differential privacy and/or support vector machines for learning in the applications with light data pattern complexities. CCS CONCEPTS• Computer systems organization → Sensor networks; • Computing methodologies → Supervised learning; • Security and privacy → Domain-specific security and privacy architectures.

PAutoBotCatcher: A blockchain-based privacy-preserving botnet detector for Internet of Things

“…In this section, we introduce the assumptions on the main entities involved in PAutoBotCatcher, namely IoT devices, gateways, block generators, and the graph aggregator. Many of them are commonly adopted by several proposals targeting the IoT scenario (e.g., [14,34,35]).…”

“…Code availability. The authors declare that they have made the source code of all the parts of PAutoBotCatcher: PAutoBotCatcher Blockchain 12 , Botnet Detection Smart Contract 13 , and (k-l)-anonymization layer 14 open source and they are available freely online.…”

Privacy-Preserving Deep Learning on Machine Learning as a Service—a Comprehensive Survey

“…The computation is limited to a polynomial of bounded degree; thus, it works in a linear nature. Another weakness of HE-based PPDL is the slow training process as it has huge complexity, and the computation process will lead to data swelling [113].…”

“…This trusted coordinator is vulnerable to a single point of failure. If this kind of failure occurs and each participant perturbs the training data, the model will yield poor accuracy [113]. Thus, a centralized coordinator is very susceptible to the single point of failure problem.…”

Recent Progress of Differentially Private Federated Learning With the Shuffle Model

“…Hence, FL can help to achieve personalization as well as enhance the performance of devices in IoT applications. In[Jiang et al (2019),Ren et al (2019)], the method is to use deep reinforcement learning to improve unloading results for IoT systems. The main idea is to consider proxy data is fewer related to the data kept on IoT machines.…”