1. Information about the Paper


Liu, Hongbo, Yu Gan, Jie Yang, Simon Sidhom, Yan Wang, Yingying Chen, and Fan Ye. “Push the limit of wifi based localization for smartphones.” InProceedings of the 18th annual international conference on Mobile computing and networking, pp. 305-316. ACM, 2012.


Highly accurate indoor localization of smartphones is critical to enable novel location based features for users and businesses. In this paper, we first conduct an empirical investigation of the suitability of WiFi localization for this purpose. We find that although reasonable accuracy can be achieved, significant errors (e.g., 6 ∼ 8m) always exist. The root cause is the existence of distinct locations with similar signatures, which is a fundamental limit of pure WiFibased methods. Inspired by high densities of smartphones in public spaces, we propose a peer assisted localization approach to eliminate such large errors. It obtains accurate acoustic ranging estimates among peer phones, then maps their locations jointly against WiFi signature map subjecting to ranging constraints. We devise techniques for fast acoustic ranging among multiple phones and build a prototype. Experiments show that it can reduce the maximum and 80-percentile errors to as small as 2m and 1m, in time no longer than the original WiFi scanning, with negligible impact on battery lifetime.


2. My Review of the Paper


The motivation behind this paper is to improve indoor localization using WiFi. The idea is that there are usually many people with smartphones in public space such as airports, shopping malls, universities. Then we can use these smartphones to help in determining more exact location using sound signal.


This paper offers new method to improve the accuracy of indoor localization using WiFi and peers’ smartphones. It is better than previous works in term that some of previous works require additional hardware or dependent to large numbers availability of access points (APs). Other contribution is that this approach does not require much battery usage.

The steps of the experiment conducted in this paper are:

  1. Initial experiment to detect the root cause of large error when using WiFi only for indoor localization; as it turns out, the root cause is that some times two faraway nodes can have similar WiFi signal strength, thus causes more errors in accurate localization. This cause can be classified into two types: (1) permanent enviromental settings such as walls, furniture placement; (2) transient factors or measurement mismatch between training and testing.
  2. Other considered factors are: orientation, holding position, number of samples, and time of the day
  3. Design peer assisted localization with following goals: peer assisted localization algorithm, concurrent acoustic ranging of multiple phones, ease of use.

Following are the steps of peer assisted localization algorithm:

  1. compute edge directions from acoustic ranging
  2. compute edge directions from initial WiFi localization
  3. graph orientation estimation
  4. movement of the graph preserving the distance variables between nodes
  5. setting the search scope
  6. joint location estimation

The beep signal is designed to be between 16 kHz – 20 kHz. It is based on previous work which concludes that cell phone microphones are more sensitive to high frequency. The other reason is that at this frequency, it will not be noticeable by other people nearby.

After evaluation, it was concluded that both peer selection and number of peers are important. But, peer selection is more important such that 3 quality peers has a good performance already. And that 10 random peers performance is relatively equal to 5 quality peers performance.



  1. This paper demonstrate more accurate way to determine a user’s location based on WiFi signal and peer acoustic ranging from smartphones around user’s location.
  2. The experiment results show that it can reduce the maximum and 80-percentile errors to as small as 2m and 1m, in time no longer than original WiFi scanning, with minimum impact on battery life.
  3. It is cited by 53 following works in just two years.


  1. Some limitations are already stated in the paper: downgrade in accuracy for more than 200 orientation, clothes and human body might attenuate the signal, errors in orientation when the target is relatively far away from the peers; when the peers are close to each other, a small orientation deviation can move the target far way from its true location while the peers are still estimated to be close to their true location.
  2. It would be better not ask the peers to send the signal manually and hence lessen their burdens. It should be done automatically, then the peers just need to confirm their willingness to help.
  3. There is latency in WiFi scanning time that the paper claims it depends on the hardware and OS of the WiFi which is outside of their control.

Other comments:

  1. The experiment conducted with same brand/type smartphones. Is there any difference or more preparation needed if we use different type of smartphone? Will this cause some effect?
  2. It might perform better with the help of machine learning algorithm.