Secure mobile cloud computing

Cloud computing is recognized as a cost-effective paradigm in outsourcing, where infrastructure, platforms, and software can be offered service. However, problems related to privacy security are still challenging, especially mobile cloud systems. end users now using systems for applications that fall under the data intensive paradigm, such skyline queries, streaming information relays, crowdsourced disaster management. above shift has opened new research directions security, interoperability, sustainable development, many other areas. With advent of fifth generation cellular telephony (5G), edge starting merge seamlessly with computing. The 5G radio access network offers some novel features, slicing, possibility operating nonpublic networks, example, within factory, but this context it may even more interesting natively supports associated each base station, core effectively defined networking supported infrastructure. This special issue presents four papers, which encompass practical approaches advance all aspects handling large amounts secure manner cloud, nodes, cloud-connected clusters MANETs wireless sensor networks. In first paper, Bartuzi Trokielewicz tackle challenge authentication devices1 through developing biometric methodology based on images human hand generated multiple light spectra. following three types devices were used recording image data: (i) visible light—a digital camera predetermined sensitivity, (ii) near-infrared—a custom-built device customized Logitech Quickcam Messenger experiments presented (iii) thermal cameras. features extracted from these then experimental analysis verification potential impact fusing different recognition thus also accuracies. Secure processing, transfer, task execution require optimization resource utilization. Such include power consumption, well communication protocols, remote deployment services. Nawrocki et al. their paper2 developed an agent-based adaptive system enabling services device, device's operation, including aspects, analyzed. They machine learning (ML) algorithms allocation tasks conventional resources. most aspect model online procedures ML methods real-time modification training sets during algorithm collected knowledge. makes approach distinctly models, process offline must updated short time intervals order simulate process. Monitoring large-scale distributed networks nodes become really challenging era Facebook, providing various e-services rapid development global trade Detection anomalies, malicious data, requires scalable centralized databases alerts confirm detect external attacks. On hand, personalization local monitoring detection tools. user redirected fake webpage or service, tricked into revealing sensitive attacker. Liu present model,3 nefarious websites avoid compromise. deployed terminals, servers, convolutional neural authors show efficiency case Low latency, ultra-high speed, high reliability important criteria IoT connectivity future smart grids (5G) potentially solve those reduction number connected sensors keeping optimal coverage reliability. Borgaonkar al.,4 pedagogical describe how grid achieve better when by utilizing mechanisms technology, offloading integral part 5G. survey state art applications, identify requirements low-voltage regulation use case. go develop threat model, discuss specifications come bear We believe papers section will serve reference students, researchers, engineers interested currently working evolving interdisciplinary area security-aware intelligent hope readers find inspiration research. grateful contributors section. thank efforts presentation recent results. would like express our sincere thanks reviewers, who have helped us ensure quality publication. Our editorial management teams Wiley's Concurrency Computation: Practice Experience journal great support throughout entire publication