Paper Summary. The paper “Spot Localization Using PHY Layer Information” analyses viability of indoor localization with application of physical layer information. Indeed, the issue of precise indoor localization has a great importance. Despite advancement of localization technology, the number of locations has significantly increased. As a result, location based applications have to meet the demands of accuracy. The paper refers to the problem of spot localization and provides an overview of different schemes that optimize distinct objectives. Finally, the paper provides an overview of indoor localization scheme, known as Precise Indoor Localization (PriLoc).
Objectives of the Paper. Localization of a target presents a complicated issue. in terms of difficulties in increasing of the accuracy of targeting. The objectives of the paper are to estimate efficiency of PinLoc and to explore the possibility of applying of PHY layer information.
Approach, Novelty, and Technical Depth. Novelty of the paper is represented by the war-driving approach. In particular, the experiment design of PinLoc is evaluated in four different environments in order to provide analysis of the accuracy of the scheme. Metrics, such as accuracy and false positives, also represent novelty. Both of them are used for evaluation of PinLoc. Additionally, the paper provides an overview of the influence of environmental changes on the PinLoc performance. Technical analysis of the PinLoc allows supporting three main hypotheses. As a result, despite random location, channel frequency responses exhibit a static structure. At the same time, the size of a particular location is small and the structure is different from other locations.
In addition, novelty of the paper is represented by evaluation approach. In particular, evaluation of Pinloc efficiency was made across 100 different spots and four environments. The novelty of the approach is also determined by appliance of two metrics, including estimating of the accuracy and testing of false positives. Thus, approach provides an opportunity not only to evaluate accuracy of the localization scheme, but to determine false positives as well. Technical depth of the paper is represented by analyzing of loclization scheme performance within different system parameters. For instance, the research discusses the influence of test packets on the PinLoc accuracy. Approach proves that even in case user demonstrates high mobility within a particular spot, localization scheme demonstrates good performance. For instance, within the environment of engineering building, PinLoc succeeded in achieving 89 percent accuracy and 7 percent of false positives. The same performance was recorded within other environments. The paper also analyses the impact of several APs, war-driving and mobility on PinLoc performance. Thus, according to the approach, performance of localization scheme can be improved with the increasing visible APs.
Major Strengths and Weaknesses. The strengths of the paper include availability of overview of main terms and principles before discussion of the results. In particular, there is an overview of data sanitization, clustering and matching modules. Moreover, the article provides the answers to the questions on application of war-driving. Evaluation of Pinloc is one of the most important strengths of the paper. For instance, evaluation was made across 100 different spots with using several test samples. In addition, localization scheme was tested in different environments, including Hudson Engineering building, student center, cafe, and museum. As a result, the paper provides an overview of effectiveness of PinLoc in different environments. Moreover, approach is based on two metrics. In particular, the first metric is represented by testing the accuracy of PinLoc. Thus, the paper examines the number of cases that require user localization on the spot. The second metric includes estimation of false positives. False positives provide that users localized to an incorrect spot. Thus, the approach analyses the possibility of occurrence of false causes when PinLoc mistakenly localizes the user. Another, strength of the paper includes availability of the comparison of PinLoc with excising localization scheme, such as Horus. Employing the approach allowed proving that the multipath signals perform stable patterns, which can demonstrate meter-scale localization. Finally, the paper offers new opportunities for localization of WiFi devices within indoor environments.
However, there are several weeaknesses of the research. In particular, not enough attention is paid to performance of PinLoc under the influence of insignificant structural changes, such as metal shelves. In addition, performance of the scheme has to be tested in accordance of the person’s movement. Every kind of motion can be sufficient for PinLoc. Results were verified on the basis of Intel 5300 hardware card. However, other cards were not considered. As a result, PinLoc sheme remains dependant on an Intel 5300 hardware card.
Learn from the Paper. Thus, according to the research, PinLoc provides 89 percent accuracy in localizing users at particular spot. In addition, application of a scheme that can precisely track user’s indoor location can improve a number of applications.
Future Work. In order to improve the approach to the issue, further development has to be done. In particular, it is necessary to develop an approach analyzing the height peculiarities and mobility of cell phones. Indeed, in the real time situations, people can carry cell phones at different heights. As a result, there is a necessity of calculating height variables. The paper also lacks investigation on the influence on PinLoc performance of the insignificant structural changes. For instance, a number of objects, such as metal shelves, can influence the accuracy of spot localization. As mentioned above, the experiments were held on the basis of Intel 5300 hardware card. However, there is a great variety of hardware cards. Thus, in order to eliminate dependency of PinLoc on a particular hardware card, further research should include testing of other cards. At the same time, the majority of available cards do not support the channel frequency responses which poses a necessity of applying another approach. In particular, in order to apply the localization scheme, the cross-platform calibration has to be done. Another area to improve is application of the approach to the 3D realities. Indeed, the results of the research were made in 2D environment. Finally, the scheme has to be applied in accordance with the MIMO. Receiver with MIMO techniques can provide a number of channel frequency responses equal to the number of antennas that receive signals. Indeed, application of multiple antennas can significantly improve spot localization.