STAMP: Enabling Privacy-Preserving Location Proofs for Mobile Users
Abstract
Introduction
Networking is the word basically relating to computers and their connectivity. It is very often used in the world of computers and their use in different connections. The term networking implies the link between two or more computers and their devices, with the vital purpose of sharing the data stored in the computers, with each other. The networks between the computing devices are very common these days due to the launch of various hardware and computer software which aid in making the activity much more convenient to build and use.
Computer security (Also known as cyber security or IT Security) is information security as applied to computers and networks. The field covers all the processes and mechanisms by which computerbased equipment, information and services are protected from unintended or unauthorized access, change or destruction. Computer security also includes protection from unplanned events and natural disasters. Otherwise, in the computer industry, the term security — or the phrase computer security — refers to techniques for ensuring that data stored in a computer cannot be read or compromised by any individuals without authorization. Most computer security measures involve data encryption and passwords. Data encryption is the translation of data into a form that is unintelligible without a deciphering mechanism. A password is a secret word or phrase that gives a user access to a particular program or system.
The explosive growth of Internet-capable and location aware mobile devices and the surge in social network usage are fostering collaborative information generation and sharing on an unprecedented scale. In particular, IDC believes that total worldwide
Smartphone shipments will reach 659.8 million units in 2012 and will grow at a CAGR of 18.6 percent until 2016.1 Almost all smartphones have cellular/ Wi-Fi Internet access and can always acquire their precise locations via pre-installed positioning software. Also owing to the growing popularity of social networks, it is more and more convenient and motivating for mobile users to share with others their experience with all kinds of points of interests (POIs) such as bars, restaurants, grocery stores, coffee shops, and hotels. Meanwhile, it becomes commonplace for people to perform various spatial POI queries at online locationbased service providers (LBSPs) such as Google and Yelp.
This paper focuses on spatial top-k queries, and the term “spatial” will be omitted hereafter for brevity. We observe two essential drawbacks with current top-k query services. First, individual LBSPs often have very small data sets comprising POI reviews. This would largely affect the usefulness and eventually hinder themore prevalent use of spatial top-k query services.
The data sets at individual LBSPs maynot cover all the Italian restaurants within a search radius. Additionally, the same restaurant may receive diverse ratings at different LBSPs, so users may get confused by very different query results from different LBSPs for the same query. A leading reason for limited data sets at individual LBSPs is that people tend to leave reviews for the same POI at one or at most only a few LBSPs’s websites which they often visit. Second, LBSPs may modify their data sets by deleting some reviews or adding fake reviews and return tailored query results in favor of the restaurants that are willing to pay or against those that refuse to pay.2 Even if LBSPs are not malicious, they may return unfaithful query results under the influence of various attacks such as the Sybil attack [2], [3] whereby the same attacker can submit many fake reviews for the same POI. In either case, top-k query users may be misled by the query results to make unwise decisions. A promising solution to the above two issues is to introduce some trusted data collectors as the central hubs for collecting POI reviews. In particular, data collectors can offer various incentives, such as free coffee coupons, for stimulating review submissions and then profit by selling the review data to individual LBSPs. Instead of submitting POI reviews to individual LBSPs, people. Similar misbehaviour has been widely reported for the websearch industry. Data contributors) can now submit them to a few data collectors to earn rewards. The data sets maintained by data collectors can thus be considered the union of the small data sets currently at individual LBSPs. Such centralized data collection also makes it much easier and feasible for data collectors to employ sophisticated defences, such as [2], [3], to filter out fake reviews from malicious entities like Sybil attackers. Data collectors can be either new service providers or more preferably existing ones with a large user base, such as Google, Yahoo, Facebook, Twitter, and MSN.
Many of these service providers (e.g., Google) have already been collecting reviews from their users and offered open APIs for exporting selected data from their systems. We postulate that they may act as location-based data collectors and sellers if sound techniques and business models are in place. The above system model is also highly beneficial for LBSPs. In particular, they no longer need struggle to solicit faithful user reviews, which is often a daunting task especially for small/medium-scale LBSPs. Instead, they can focus their limited resources on developing appealing functionalities (such as driving directions and aerial photos) combined with the high-quality review data purchased from data collectors. The query results they can provide will be much more trustworthy, which would in turn help them attract more and more users. This system model thus can greatly help lower the entrance bar for new LBSPs without sufficient funding and thus foster the prosperity of location- based services and applications. A main challenge for realizing the appealing system above is how to deal with untrusted and possibly malicious LBSPs. Specifically, malicious LBSPs may still modify the data sets from data collectors and provide biased top-k query results in favor of POIs willing to pay. Even worse, they may falsely claim generating query results based on the review data from trusted data collectors which they actually did not purchase. Moreover, nonmalicious LBSPs may be compromised to return fake top-k query results.
In this paper, we propose three novel schemes to tackle the above challenge for fostering the practical deployment and wide use of the envisioned system. The key idea of our schemes is that the data collector precomputes and authenticates some auxiliary information (called authenticated hints) about its data set, which will be sold along with its data set to LBSPs. To faithfully answer a top-k query, a LBSP need return the correct top-k POI data records as well as proper authenticity and correctness proofs constructed from authenticated hints. The authenticity proof allows the query user to confirm that the query result only consists of authentic data records from the trusted data collector’s data set, and the correctness proof enables the user to verify that the returned top-k POIs are the true ones satisfying the query. The first two schemes both target snapshot top-k queries but differ in how authenticated hints are precomputed and how authenticity and correctness proofs are constructed and verified as well as the related communication and computation overhead. The third scheme, built upon the first scheme, realizes efficient and verifiable moving top-k queries. The efficacy and efficiency of our schemes are thoroughly analyzed and evaluated through detailed simulation studies
Existing schemes which require multiple trusted or semi-trusted third parties, STAMP requires only Single semi-trusted third party which can be embedded in a Certificate Authority (CA). We design our system with an objective of protecting users’ anonymity and location privacy. No parties other than verifiers could see both a user’s identity and STP information (verifiers need both identity and STP information in order to perform verification and provide services). Users are given the flexibility to choose the location granularity level that is revealed to the verifier. We examine two type s of collusion attacks: (1) A user who is at an intended location masquerades s another colluding user and obtains STP proofs for . This attack has never been addressed in any existing STP proof schemes. (2) Colluding users mutually generate fake STP proofs for each other. There have been efforts to address this type of collusion. However, existing solutions suffer from high computational cost and low scalability. Particularly, the latter collusion scenario is in fact the challenging Terrorist Fraud attack, which is a critical issue for our targeted system, but none of the existing systems has addressed it. We integrate the Bussard-Bagga distance bounding protocol into STAMP to protect our scheme against this collusion attack. Collusion scenario (1) is hard to prevent without a trusted third party. To make our system resilient to this attack, we propose an entropy-based trust model to detect the collusion scenario. We implemented STAMP on the Android platform and carried out extensive validation experiments. The experimental results show that STAMP requires low computational overhead.
Proposed system
In this paper, we propose an STP proof scheme named Spatial-Temporal provenance Assurance with Mutual Proofs (STAMP). STAMP aims at ensuring the integrity and non-transferability of the STP proofs, with the capability of protecting users’ privacy. Most of the existing STP proof schemes rely on wireless infrastructure (e.g., WiFi APs) to create proofs for mobile users. However, it may not be feasible for all types of applications, e.g., STP proofs for the green commuting and battlefield examples certainly cannot be obtained from wireless APs. To target a wider range of applications, STAMP is based on a distributed architecture. Co-located mobile devices mutually generate and endorse STP proofs for each other, while at the same time it does not eliminate the possibility of utilizing wireless infrastructures as more trusted proof generation sources. In addition, in contrast to most of the existing schemes which require multiple trusted or semi-trusted third parties, STAMP requires only a single semi-trusted third party which can be embedded in a Certificate Authority (CA). We design our system with an objective of protecting users’ anonymity and location privacy. No parties other than verifiers could see both a user’s identity and STP information (verifiers need both identity and STP information in order to perform verification and provide services). Users are given the flexibility to choose the location granularity level that is revealed to the verifier.
Conclusion
In this STAMP: Enabling Privacy-Preserving Location Proofs for Mobile Users paper we have presented STAMP, which aims at providing security and privacy assurance to mobile users’ proofs for their past location visits. STAMP relies on mobile devices in vicinity to mutually generate location proofs or uses wireless AP sto generate location proofs.Integritya nd non-transferability of location proofs and location privacy of users are the main design goals of STAMP. We have specifically dealt with two collusion scenarios: P-P collusion and P-W collusion. To protect against P-P collusions, we integrated the Bussard-Bagga distance bounding protocol into the design of STAMP. To detect P-W collusion, we proposed an entropy-based trustmodelto evaluate the trust levelofclaimsof the pastlocation visits.Oursecurity analysis shows that STAMP achieves the security and privacy objectives. Our implementation on Android smartphones indicates that low computational and storage resources are required to execute STAMP. Extensive simulation results show that our trust model is able to attain a high balanced accuracy with appropriate choices of system parameters.
