Bharadwaj  Veeravalli

Prof. Bharadwaj Veeravalli

National University of Singapore, Singapore City, Singapore

Research Focus & Metrics

Academic Career Length

24Years (2003-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 March 2019 in IEEE Transactions on Emerging Topics in Computing
Verified
Journal Article
Closed access
ABAhmad Qaisar Al Badawi
Ahmad Qaisar Al Badawi
YPYuriy Polyakov
Yuriy Polyakov
KAKhin Mi Mi Aung
Khin Mi Mi Aung
BVBharadwaj Veeravalli
Bharadwaj Veeravalli
KRKurt Rohloff
Kurt Rohloff
Homomorphic encryption is an emerging form of encryption that provides the ability to compute on encrypted data without ever decrypting them. Potential applications include aggregating sensitive encrypted data on a cloud environment and computing on the data in the cloud without compromising data privacy. There have been several recent advances resulting in new homomorphic encryption schemes and optimized variants. We implement and evaluate the performance of two optimized variants, namely Bajard-Eynard-Hasan-Zucca (BEHZ) and Halevi-Polyakov-Shoup (HPS), of the most promising homomorphic encryption scheme in CPU and GPU. The most interesting (and also unexpected) result of our performance evaluation is that the HPS variant in practice scales significantly better (typically by 15%-30%) with increase in multiplicative depth of the computation circuit than BEHZ, implying that the HPS variant will always outperform BEHZ for most practical applications. For the multiplicative depth of 98, our fastest GPU implementation performs homomorphic multiplication in 51 ms for 128-bit security settings, which is faster by two orders of magnitude than prior results and already practical for cloud environments supporting GPU computations. Large multiplicative depths supported by our implementations are required for applications involving deep neural networks, logistic regression learning, and other important machine learning problems.

18 September 2019 in IEEE Sensors Journal
Verified
Journal Article
Closed access
DWDavid Liang Tai Wong
David Liang Tai Wong
JYJufeng Yu
Jufeng Yu
YLYongfu Li
Yongfu Li
CDChacko John Deepu
Chacko John Deepu
DNDuy Hoa Ngo
Duy Hoa Ngo
CHChun-Huat Heng
Chun-Huat Heng
A compact, light-weight and low-power wireless vital signs monitoring system based on Wireless Body Area Network (WBAN) protocol has been developed. The system, VySys, includes two compact wearable wireless biosensor devices for continuous vital signs capturing and transmission, a gateway to relay the message collected from the biosensors to cloud, and finally a client apps to access and display the stored data in the cloud. Both biosensor devices can last for 24 hours and weigh less than 22 g and 44 g, respectively. They consist of proprietary in-house bio-sensing integrated circuit (IC) and commercial off-the-shelf Bluetooth Low Energy (BLE) module. VySys has been deployed in clinical trials with 14 subjects. From the studies, the accuracy and advantage of VySys are evaluated and the five vital signs captured (heart rate (HR), respiration rate (RR), temperature (TMP), oxygen saturation (SpO2) and systolic blood pressure (SBP)) are benchmarked against a commercial medical-grade device. The results show strong statistical correlation (r > 0.68). In terms of clinical significance, all its mean difference are within limits of accepted clinical discrepancies. In terms of efficacy by comparing against the best known reported results, (1) VsSys is more precise by 28.2%, 36.2%, 70.0%, 37.6% and 34.4% for HR, RR, TMP, SpO2 and SBP, respectively and (2) has a narrower 95% limit of agreement (LoA) by 24.5%, 23.9%, 50.6%, 37.4% and 34.4% for HR, RR, TMP, SpO2 and SBP, respectively.

8 October 2019 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
RCRonghui Cao
Ronghui Cao
ZTZhuo Tang
Zhuo Tang
CLChubo Liu
Chubo Liu
BVBharadwaj Veeravalli
Bharadwaj Veeravalli
Nowadays, Cloud-supported Internet of Things (Cloud-IoT) has been broadly deployed in smart medical systems, where the limitations of IoT associated medical devices in terms of data access, storage, scalability and computing are solved through the use of cloud computing architectures. However, with the rapid development of medical equipment and the increasing number of medical devices, it will be extremely difficult to program or manage such an expanding and massive medical IoT system in traditional single-cloud platforms. In this paper, we design and implement a multi-cloud framework for building OpenStack-based platform for medical IoT, referred to as Tri-Storage Failure Recovery System (Tri-SFRS). To implement Tri-SFRS, we combine several techniques to achieve this reduction in effort, including a multi-cloud cascading architecture, a low-overhead native testing framework, a medical data storage-backup mechanism, and snapshot-volume cascaded operations for b-ultrasonic data. Tri-SFRS is also able to simultaneously enable resource management specialization. Tri-SFRS has been designed as a native component in the OpenStack platform, and it demonstrates the degree of native OpenStack multi-cloud platform management by our proposed cascading framework. Comparing with the traditional single-cloud OpenStack platform, Tri-SFRS can reduce the resource-request processing latency from B ultrasonic machines by up to 20%. Our experiments also demonstrate the broad applicability of Tri-SFRS.

2 September 2020 in IEEE Transactions on Parallel and Distributed Systems
Verified
Journal Article
Closed access
ABAhmad Qaisar Al Badawi
Ahmad Qaisar Al Badawi
BVBharadwaj Veeravalli
Bharadwaj Veeravalli
JLJie Lin
Jie Lin
NXNan Xiao
Nan Xiao
MKMatsumura Kazuaki
Matsumura Kazuaki
AQAhmad Qaisarahmadalbadawi
Ahmad Qaisarahmadalbadawi
We present a multi-GPU design, implementation and performance evaluation of the Halevi-Polyakov-Shoup (HPS) variant of the Fan-Vercauteren (FV) levelled Fully Homomorphic Encryption (FHE) scheme. Our design follows a data parallelism approach and uses partitioning methods to distribute the workload in FV primitives evenly across available GPUs. The design is put to address space and runtime requirements of FHE computations. It is also suitable for distributed-memory architectures, and includes efficient GPU-to-GPU data exchange protocols. Moreover, it is user-friendly as user intervention is not required for task decomposition, scheduling or load balancing. We implement and evaluate the performance of our design on 2 homogeneous and heterogeneous NVIDIA GPU clusters: K80, and a customized P100. We also provide a comparison with a recent shared-memory based multi-core CPU implementation using two homomorphic circuits as workloads: vector addition and multiplication. Moreover, we use our multi-GPU Levelled-FHE to implement the inference circuit of two Convolutional Neural Networks (CNNs) to perform homomorphically image classification on encrypted images from the MNIST and CIFAR-10 datasets. Our implementation provides 1 to 3 orders of magnitude speedup compared with the CPU implementation on vector operations. In terms of scalability, our design shows reasonable scalability curves when the GPUs are fully connected.

11 September 2023 in IEEE Transactions on Parallel and Distributed Systems
Verified
Journal Article
Closed access
YHYi Hu
Yi Hu
HWHao Wang
Hao Wang
LWLiangyuan Wang
Liangyuan Wang
MHMenglan Hu
Menglan Hu
KPKai Peng
Kai Peng
BVBharadwaj Veeravalli
Bharadwaj Veeravalli
Microservices are an architectural and organizational paradigm for Internet application development. In cloud data centers, delay-sensitive applications receive massive user requests, which are fed into multiple queues and subsequently served by multiple microservice instances. Accordingly, effective deployment of multiple queues and containers significantly reduces queuing delay, processing delay, and communication delay, boosting overall service performance. Due to the increased complexity of call dependencies, probabilistic routing paths, and microservice multiplexing, the deployment of service instances fully interacts with request routing, bringing great difficulties to service orchestration. In this case, it is valuable to simultaneously consider service deployment and request routing in a fine-grained manner. However, most existing studies considered them as two independent components with local optimization, while data dependencies and the deployment of multiple queues/containers are not well handled. Therefore, this paper proposes to jointly optimize the deployment and request routing of microservice call graphs based on fine-grained queuing network analysis and container orchestration. We first formulate the joint optimization problem as a mixed-integer nonlinear program and then exploit open Jackson queuing networks to model the intrinsic data dependencies and analyze response latency and microservice multiplexing. To optimize the overall cost and latency, this paper presents an efficient two-stage heuristic algorithm. The algorithm consists of a resource-splitting-based deployment approach and a partition-mapping-based routing method, which employ instance-level optimization to enhance performance. Further, this paper also provides mathematical analysis on the performance and complexity of the propose algorithm. Finally, comprehensive trace-driven experiments demonstrate that the overall performance of our approach is better than existing microservice benchmarks. Extensive simulation results show that average deployment cost is reduced by 27.4% and the end-to-end response latency is reduced by 15.1% on average.

5 October 2017 in Multimedia Tools and Applications
Verified
Journal Article
Closed access
PDF file
QXQuanqing Xu
Quanqing Xu
CJChao Jin
Chao Jin
MRMohamed Faruq Bin Mohamed Rasid
Mohamed Faruq Bin Mohamed Rasid
BVBharadwaj Veeravalli
Bharadwaj Veeravalli
KAKhin Mi Mi Aung
Khin Mi Mi Aung