1304 IEEE COMMUNICATIONS SURVEYS TUTORIALS.VOL.16.NO.3.THIRD QUARTER 2014 TABLE VI etc.The importance of these indicators depends on the partic- CHALLENGES AND OPPORTUNITIES FOR AUTHENTICATION AND PRIVACY ular application requirements. PROTECTION Generally speaking,RFID-based localization can be classi- Mechanism Challenges Opportunities fied into tag localization and reader localization.In the tag Physical Implementing in a recy The signal spectral feature localization,each object to be located is attached with an mechanism cled and delicate approach in physical layer can be instead of brute-force ap- used to mitigate comput- RFID tag and one or more RFID readers are deployed in the proach. ing resource requirement. environment.A server gathers data from the readers,executes Symmetric Complex operations and Lightweight encryption a localization algorithm and notifies the result to the object. key high requirement for com- and fast searching over encryption puting resources. encrypted data can be In the reader localization,the object carries an RFID reader used to complement the and a set of RFID tags are deployed in the environment.The protocols object uses the reader to actively obtain its own location.In the Hash func- Encrvotion and decrvoton Forward security can tion cannot be effectively im- be cleverly leveraged following,we respectively introduce the principles and related plemented. for lightweight work in the tag localization and reader localization. authentication. B.Locate the Tags V.LOCALIZATION AND ACTIVITY SENSING In regard to the tag localization,the tag is usually located In a number of RFID-based applications,the users conven- according to the Received Signal Strength Indicator (RSSD). tionally have more demands than simply identifying the RFID A straightforward approach to use RSSI is building a model tags.They require not only to"identify"the tags but also to to depict the mapping relationship between the RSSI and "locate"the tags.The properties of the RFID tags like the re- the distance.In this way,the actual distance between the ceived signal strength(RSS)have brought new opportunities to tag and the reader can be measured by the received signal conduct accurate localization.For example,the backscattered strength in the reader.Moreover,the measured distances from signal strength has a significant reverse relationship with the an unknown tag to several RFID readers constrain the presence distance between the reader and the tag.Generally speaking, of this tag.The exact location of this tag can be effectively as the distance increases,the backscattered signal strength estimated by using the method of trilateration or multilatera- is monotonically decreasing.Moreover,due to the far field tion.Therefore,for the distance-based positioning.the most propagation and the backscatter property,the signal strength important thing is how to measure distances in the physical is rather sensitive to the distance in the limited scanning range. world.In theory,radio signal strengths diminish with distance Therefore,it is possible to measure the distance according to according to a power law.A generally employed model for the backscattered signal and further estimate the location.By wireless radio propagation is proposed in [65].However,the effectively leveraging the above information,the RFID system radio propagation in RFID systems is severely impacted by is able to accurately locate the objects. the issues like ambient noises,path loss and multi-path effect, Furthermore,it is noted that the backscattered signal which makes the above model rather unreliable.How to strength is also very sensitive to the surrounding environment, eliminate such effect so as to enhance the indoor localization i.e.,the RSSI for a tag changes significantly when an object performance is a big challenge.Many ranging techniques are (e.g.,a person)is passing by it.Therefore,a tag-free RFID- proposed and developed in RFID systems.Hightower et al. based activity sensing is inspired from the above phenomenon. present SpotON [66],a new tagging technology for 3D local- In this section,we respectively introduce the research progress ization based on radio signal strength analysis.They propose in localization and activity sensing based on RFID technology. an aggregation algorithm to minimize signal strength error relative to empirical data.After mapping the signal strength A.Accurate Localization measurements to an approximate distance,they aggregate the values to triangulate the precise position of the tagged object. In regard to the localization,there are three key performance Most existing indoor positioning systems can't adapt to the indicators as follows: environmental variations well,as they need an accurate signal Accuracy:the object should be accurately located to propagation model.Xiao et al.propose an environmental- satisfy the demand for context-aware or location-based adaptive RSSI-based indoor positioning approach using RFID services in pervasive applications,i.e.,the estimation [67],in which the parameters of signal propagation model are error for localization should be made as small as possible. updated online in a closed-loop feedback correction manner. Time-efficiency:if the object is continuously moving,it Brchan et al.propose a real-time localization system [68]. is crucially important to locate the object in a real-time using efficient multiple propagation models to compensate for approach,i.e.,the localization should be executed within the drawback of the received signal strength technique. specified time delay. Another approach to use RSSI is collecting the fingerprints Cost:the localization conventionally requires some pre- of the measured RSSI at each location of the positioning cision instruments to measure the critical parameters, area.Suppose multiple readers are deployed in the positioning therefore,a cost-effective localization system is usually area,when a tag is deployed at the labeled location,the RSSI more preferred. fingerprint can be collected by measuring the received signal Besides,there are a number of performance indicators for the strength at each reader.In this way,the RSSI fingerprint localization,including the fault tolerance,energy-efficiency, is recorded as a vector,with its size equal to the number1304 IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 16, NO. 3, THIRD QUARTER 2014 TABLE VI CHALLENGES AND OPPORTUNITIES FOR AUTHENTICATION AND PRIVACY PROTECTION Mechanism Challenges Opportunities Physical mechanism Implementing in a recy￾cled and delicate approach instead of brute-force ap￾proach. The signal spectral feature in physical layer can be used to mitigate comput￾ing resource requirement. Symmetric￾key encryption Complex operations and high requirement for com￾puting resources. Lightweight encryption and fast searching over encrypted data can be used to complement the protocols. Hash func￾tion Encryption and decryption cannot be effectively im￾plemented. Forward security can be cleverly leveraged for lightweight authentication. V. LOCALIZATION AND ACTIVITY SENSING In a number of RFID-based applications, the users conven￾tionally have more demands than simply identifying the RFID tags. They require not only to “identify” the tags but also to “locate” the tags. The properties of the RFID tags like the re￾ceived signal strength (RSS) have brought new opportunities to conduct accurate localization. For example, the backscattered signal strength has a significant reverse relationship with the distance between the reader and the tag. Generally speaking, as the distance increases, the backscattered signal strength is monotonically decreasing. Moreover, due to the far field propagation and the backscatter property, the signal strength is rather sensitive to the distance in the limited scanning range. Therefore, it is possible to measure the distance according to the backscattered signal and further estimate the location. By effectively leveraging the above information, the RFID system is able to accurately locate the objects. Furthermore, it is noted that the backscattered signal strength is also very sensitive to the surrounding environment, i.e., the RSSI for a tag changes significantly when an object (e.g., a person) is passing by it. Therefore, a tag-free RFID￾based activity sensing is inspired from the above phenomenon. In this section, we respectively introduce the research progress in localization and activity sensing based on RFID technology. A. Accurate Localization In regard to the localization, there are three key performance indicators as follows: • Accuracy: the object should be accurately located to satisfy the demand for context-aware or location-based services in pervasive applications, i.e., the estimation error for localization should be made as small as possible. • Time-efficiency: if the object is continuously moving, it is crucially important to locate the object in a real-time approach, i.e., the localization should be executed within specified time delay. • Cost: the localization conventionally requires some pre￾cision instruments to measure the critical parameters, therefore, a cost-effective localization system is usually more preferred. Besides, there are a number of performance indicators for the localization, including the fault tolerance, energy-efficiency, etc. The importance of these indicators depends on the partic￾ular application requirements. Generally speaking, RFID-based localization can be classi- fied into tag localization and reader localization. In the tag localization, each object to be located is attached with an RFID tag and one or more RFID readers are deployed in the environment. A server gathers data from the readers, executes a localization algorithm and notifies the result to the object. In the reader localization, the object carries an RFID reader and a set of RFID tags are deployed in the environment. The object uses the reader to actively obtain its own location. In the following, we respectively introduce the principles and related work in the tag localization and reader localization. B. Locate the Tags In regard to the tag localization, the tag is usually located according to the Received Signal Strength Indicator (RSSI). A straightforward approach to use RSSI is building a model to depict the mapping relationship between the RSSI and the distance. In this way, the actual distance between the tag and the reader can be measured by the received signal strength in the reader. Moreover, the measured distances from an unknown tag to several RFID readers constrain the presence of this tag. The exact location of this tag can be effectively estimated by using the method of trilateration or multilatera￾tion. Therefore, for the distance-based positioning, the most important thing is how to measure distances in the physical world. In theory, radio signal strengths diminish with distance according to a power law. A generally employed model for wireless radio propagation is proposed in [65]. However, the radio propagation in RFID systems is severely impacted by the issues like ambient noises, path loss and multi-path effect, which makes the above model rather unreliable. How to eliminate such effect so as to enhance the indoor localization performance is a big challenge. Many ranging techniques are proposed and developed in RFID systems. Hightower et al. present SpotON [66], a new tagging technology for 3D local￾ization based on radio signal strength analysis. They propose an aggregation algorithm to minimize signal strength error relative to empirical data. After mapping the signal strength measurements to an approximate distance, they aggregate the values to triangulate the precise position of the tagged object. Most existing indoor positioning systems can’t adapt to the environmental variations well, as they need an accurate signal propagation model. Xiao et al. propose an environmental￾adaptive RSSI-based indoor positioning approach using RFID [67], in which the parameters of signal propagation model are updated online in a closed-loop feedback correction manner. Brchan et al. propose a real-time localization system [68], using efficient multiple propagation models to compensate for the drawback of the received signal strength technique. Another approach to use RSSI is collecting the fingerprints of the measured RSSI at each location of the positioning area. Suppose multiple readers are deployed in the positioning area, when a tag is deployed at the labeled location, the RSSI fingerprint can be collected by measuring the received signal strength at each reader. In this way, the RSSI fingerprint is recorded as a vector, with its size equal to the number