Vehicular networks enable vehicles to generate and broadcast messages in order improve traffic safety efficiency. However, due the nontrusted environments, it is difficult for evaluate credibilities of received messages. In this paper, we propose a decentralized trust management system vehicular based on blockchain techniques. system, can validate from neighboring using Bayesian Inference Model. Based validation result, vehicle will rating each message source vehicle. With ratings uploaded...

In cloud computing, data owners host their on servers and users (data consumers) can access the from servers. Due to outsourcing, however, this new paradigm of hosting service also introduces security challenges, which requires an independent auditing check integrity in cloud. Some existing remote checking methods only serve for static archive and, thus, cannot be applied since dynamically updated. Thus, efficient secure dynamic protocol is desired convince that are correctly stored paper,...

With the flourishing and advancement of IoT, smart city has become an emerging paradigm, consisting ubiquitous sensing, heterogeneous network infrastructure, intelligent information processing control systems. A can monitor physical world in real time, provide services to both local residents travelers terms transportation, healthcare, environment, entertainment, energy. However, security privacy concerns arise, since applications not only collect a wide range privacy-sensitive from people...

As an emerging training model with neural networks, federated learning has received widespread attention due to its ability update parameters without collecting users' raw data. However, since adversaries can track and derive participants' privacy from the shared gradients, is still exposed various security threats. In this paper, we consider two major issues in process over deep networks (DNNs): 1) how protect user's (i.e., local gradients) 2) verify integrity (or correctness) of aggregated...

Big data offers a plethora of opportunities to mobile network operators for improving quality service. This article explores various means integrating big analytics with optimization toward the objective user experience. We first propose framework Data-Driven (BDD) optimization. then present characteristics that are collected not only from equipment but also networks. Moreover, several techniques in collection and discussed viewpoint Certain cases on application proposed performance given...

Data access control is an effective way to ensure data security in the cloud. However, due outsourcing and untrusted cloud servers, becomes a challenging issue storage systems. Existing schemes are no longer applicable systems, because they either produce multiple encrypted copies of same or require fully trusted server. Ciphertext-policy attribute-based encryption (CP-ABE) promising technique for data. inefficiency decryption revocation, existing CP-ABE cannot be directly applied construct...

Data access control is an effective way to ensure the data security in cloud. Due outsourcing and untrusted cloud servers, becomes a challenging issue storage systems. Ciphertext-Policy Attribute-based Encryption (CP-ABE) regarded as one of most suitable technologies for storage, because it gives owners more direct on policies. However, difficult directly apply existing CP-ABE schemes systems attribute revocation problem. In this paper, we design expressive, efficient revocable scheme...

In this paper, we analyze the inherent characteristic of electronic medical records (EMRs) from actual health (eHealth) systems, where found that first, multiple patients would generate large amounts duplicate EMRs and second, cross-patient be generated numerously only in case consult doctors same department. We then propose first efficient secure encrypted deduplication scheme for cloud-assisted eHealth systems (HealthDep). With integration our analysis results, HealthDep allows cloud...

The mobile crowdsourcing network (MCN) is a promising architecture that applies the principles of to perform tasks with human involvement and powerful devices. However, it also raises some critical security privacy issues impede application MCNs. In this article, in order better understand these challenges, we first propose general for comprising both sensing computing. After that, set forth several challenges essentially capture characteristics We formulate research problems leading...

As a basic query function, range has been exploited in many scenarios such as SQL retrieves, location-based services, and computational geometry. Meanwhile, with explosive growth of data volume, users are increasingly inclining to store on the cloud for saving local storage cost. However, long-standing problem is that user's may be completely revealed server because it full access right. To cope this problem, frequently-used method encrypt raw before outsourcing them, but availability...

Many artificial intelligence technologies have been applied for extracting useful information from massive industrial big data. However, the privacy issues are usually overlooked in many existing methods. In this article, we propose an anonymous and privacy-preserving federated learning scheme mining of We explored effect proportion shared parameters on accuracy through experiments, found that sharing partial can almost achieve all parameters. On basis, our proposed reduces leakage by fewer...

The security of vehicular networks has been paid increasing attention to with the rapid development automobile industry and Internet Things (IoT). However, existing approaches mainly focus on ensuring data authentication integrity, which are not sufficient assess credibility received messages. Recently, reputation systems proved be effective solve above problem. This paper proposes a new system for assessment based blockchain techniques. In this system, vehicles rate messages observations...

With the flourishing of multi-functional wearable devices and widespread use smartphones, MHN becomes a promising paradigm ubiquitous healthcare to continuously monitor our health conditions, remotely diagnose phenomena, share information in real time. However, MHNs raise critical security privacy issues, since highly sensitive is collected, users have diverse requirements about such information. In this article, we investigate protection from perspective QoP, which offers adjustable...

Crowdsourcing is a distributed computing paradigm that utilizes human intelligence or resources from crowd of workers. Existing solutions task recommendation in crowdsourcing may leak private and sensitive information about both tasks To protect privacy, workers should be encrypted before being outsourced to the platform, which makes challenging problem. In this paper, we propose privacy-preserving scheme (PPTR) for crowdsourcing, achieves task-worker matching while preserving privacy worker...

A cloud storage service allows data owner to outsource their the and through which provide access users. Because server are not in same trust domain, semi-trusted cannot be relied enforce policy. To address this challenge, traditional methods usually require encrypt deliver decryption keys authorized These methods, however, normally involve complicated key management high overhead on owner. In paper, we design an control framework for systems that achieves fine-grained based adapted...

How to control the access of huge amount big data becomes a very challenging issue, especially when are stored in cloud. Ciphertext-policy attribute-based encryption (CP-ABE) is promising technique that enables end-users encrypt their under policies defined over some attributes consumers and only allows whose satisfy decrypt data. In CP-ABE, policy attached ciphertext plaintext form, which may also leak private information about end-users. Existing methods partially hide attribute values...

Data access control is an effective way to ensure the data security in cloud. However, due outsourcing and untrusted cloud servers, becomes a challenging issue storage systems. Existing schemes are no longer applicable systems, because they either produce multiple encrypted copies of same or require fully trusted server. Ciphertext-Policy Attribute-based Encryption (CP-ABE) promising technique for data. It requires authority manages all attributes distributes keys system. In there...

IoT is envisioned as the next stage of information revolution, enabling various daily applications and providing better service by conducting a deep fusion with cloud fog computing. As key mission most applications, data management, especially fundamental function-data query, has long been plagued severe security privacy problems. Most query providers, including big ones (e.g., Google, Facebook, Amazon, so on) are suffering from intensive attacks launched insiders or outsiders. consequence,...

Due to the high volume and velocity of big data, it is an effective option store data in cloud, as cloud has capabilities storing processing user access requests. Attribute-based encryption (ABE) a promising technique ensure end-to-end security cloud. However, policy updating always been challenging issue when ABE used construct control schemes. A trivial implementation let owners retrieve re-encrypt under new policy, then send back This method, however, incurs communication overhead heavy...

Benefiting from the advancement of algorithms in massive data and powerful computing resources, deep learning has been explored a wide variety fields produced unparalleled performance results. It plays vital role daily applications is also subtly changing rules, habits, behaviors society. However, inevitably, data-based strategies are bound to cause potential security privacy threats, arouse public as well government concerns about its promotion real world. In this article, we mainly focus...

With the ever-increasing demands on multimedia applications, cloud computing, due to its economical but powerful resources, is becoming a natural platform process, store, and share contents. However, employment of computing also brings new security privacy issues as few public servers can be fully trusted by users. In this paper, we focus how securely video contents certain group people during particular time period in cloud-based systems, propose cryptographic approach, provably secure...

As a very attractive computing paradigm, cloud makes it possible for resource-constrained users to enjoy cost-effective and flexible resources of diversity. Considering the untrustworthiness servers data privacy users, is necessary encrypt before outsourcing cloud. However, form encrypted storage also poses series problems, such as: How can search over outsourced data? realize user-side verifiability results resist malicious servers? enable server-side check owners? achieve payment fairness...

Password-based single-sign-on authentication has been widely applied in mobile environments. It enables an identity server to issue tokens users holding correct passwords. With token, one can request services from related service providers without multiple registrations. However, if adversary compromises the server, he retrieve users' passwords by performing dictionary guessing attacks (DGA) and overissue break security. In this paper, we propose a password-based threshold scheme dubbed...

Secure and efficient access authentication is one of the most important security requirements for vehicular networks, but it difficult to fulfill due potential attacks long delay caused by high vehicle mobility, etc. Most existing protocols, either do not consider like single point failure or focus on reducing delay. To address these issues, we introduce an edge-assisted decentralized (EADA) architecture, which provides secure more communication-efficient enabling server delegate its...