Smartphone used as Antenna to Accurate Outdoor AP Location --Summary of “I am the Antenna: Accurate Outdoor AP Location Using Smartphones” 1. Background WiFi networks are universally used in our everyday lives. It provides wireless connectivity for a wide range of mobile networking devices. All digital devices can connect to the internet via a wireless network access point that has a range of about 20 meters indoors such as homes and offices and a greater range outdoors such as parks, schools and shopping centers.As Internet users become reliant on WiFi access points (APs) to connect their smartphones, tablets and laptops, the availability and performance of tomorrow’s networks will depend on well tuned and managed access points. A key part of managing access points is the ability to locate individual access points based on their signal.
Current techniques (RSS Grandient, Signal Map, Directional Antenna) to locate outdoor WiFi access points require extensive wardriving measurements and low accuracy, determined by significant offline computation or complex hardware components that would cost several thousand dollars.In order to solve this problem, there is a commercial need for a cost and time-efficient alternative way to accurate outdoor WiFi AP location. 2. A key insight for Accurate AP location A potential solution that focus on cost and time-efficient using common off-the-shelf hardware would make it available to home users and small business managing their local hotspots. So we propose a way to locate APs in real-time using smartphones.
Our insight is that by rotating a standard wireless receiver (smartphone) around a blocking object, we can effectively emulate the sensitivity and functionality of a directional antenna.By “rotating” the receiver’s position with respect to the obstacle, and observing the received signal strength, we can determine the approximate direction of the transmitter. This process can be recognized as directional analysis. We assume that a user can accurately locate WiFi APs using common-off-the-shelf smartphones as receivers, and her own body as the blocking obstacle. To perform a directional analysis operation, she slowly rotates her body around 360 degrees, while keeping the smartphone in front of her and performing periodic received signal strength (RSS) measurements.
The observed RSS should be at its lowest point when the user’s body is directly between the smartphone and the wireless AP. The hypothesis has extensive feasibility that we can detect these signal strength artifacts on different kinds of smartphone platforms (e. g. : Android, Windows mobile, Apple Iphone4) for a variety of outdoor environments (e. g.
: Simple Line-of-Sight, Complex LOS, No LOS ). The main idea of insight is that we can develop a model for detecting signal dips caused by blocking obstacles, and use it to produce a directional analysis technique that accurately predicts the direction of AP with an associated confidence value. . Blocking Obstacle Effect According to mentioned above, it’s necessary to describe our simple problem scene that a user who hold a smartphone would like to find the physical location of a WiFi AP through its BSSID, with the purpose of locating a transmitting AP rather than determining his own location.
We focus on Received Signal Strength (RSS) as the physical modality to locate a transmitter. We apply the insight for a consumption on our context of smartphone based AP location. It’s obvious the body of a user holding a smartphone will block a portion of incoming WiFi signal.The closer the user is to being on the straight line between the smartphone and the AP, the weaker the signal perceived by the phone. This effect of human body has been observed based on a variety of frequencies and radio hardware, including indoor environments as well. Figure1 Figure1 shows that when facing the AP (1), the body is not an obstacle while on the other side, his back is towards the AP (2), the user’s body becomes an obstacle and blocks the signals from reaching the smartphone.
We make the user rotates himself in place (c), the received signal on phone displays an interesting phenomenon, there will be a peak when he faces the AP and a dip when his back faces the AP. So a user can gain a hint of which direction points towards the AP by measuring signal strength at different rotational angles. There are detailed smartphone experiments that not only confirm the assumption, but also study the impact the body act as an obstacle has the signal on strength as the user rotates himself.The experiments also help to understand the impact on results based on different causes such as propagation environment (LOS, Complex LOS, No LOS), phone hardware (Droid, G1, IPhone, Windows Mobile) and WiFi standards (802. 11b/g, 802. 11n).
From the experiments results we can get the two key observations (Shown as Figure2). First, the position of the user’s body can importantly affect the smartphone’s received WiFi signal strength. When the user makes his back facing the AP, His body becomes an obstacle and degrades the received signal strength.Second, we use BF (Back facing AP) to mark the opposite AP direction where the blocking effect should be at its strongest. And it’s clear that an accurate AP location system can’t simply rely on finding the angle with lowest signal strength, should use more sophisticated techniques to indicate the AP direction. Figure2.
Two observations 4. Accurate Access Point location using Borealis System According to our observations, we develop Borealis , a new AP localization system for commodity smartphones that makes signal strength artifacts to compute the direction of an AP.Borealis uses off-the-shelf smartphones and produces real-time results with a few of measurements. And Borealis users can perform robust directional analysis by turning their bodies on a 360°axis.
Borealis has two critical requirements. First, Borealis must use minimal energy and computational resource in its directional analysis since the resources on smartphones is limited. Second, Borealis must produce results in real-time, and its directional measurements should be minimized to save user effort.We can meet both of the goals by deploying a system that minimizes computation, the more accurate the result, the fewer number of direction analysis measurements are necessary, so saving both device battery and user effort.
The biggest challenge of Borealis is that we cannot take the angle of lowest signal strength to estimate AP direction because it will lead to large estimation errors. There are two factors causing this. First, we measure the signal strength of various directions sequentially but not simultaneously as the user rotates. Second, there will be measurements noise since the measurements time duration at each angle is limited.
We define the range of signal strength degrading heavily during user rotation as the blocking sector. The angular size of the sector is roughly 90°under general configurations. We propose to estimate AP direction by locating the blocking sector within the signal strength profile. The opposite direction from the center of the blocking sector is the AP direction. Borealis’ directional analysis also provides a confidence level associated with each estimate of AP direction.
The confidence value is computed by comparing the signal strength profile to an ideal profile derived from our abstract model on body blocking.So we can use the confidence to capture the similarity between the measured signal strength profile and the ideal profile with the purpose of controlling how often a user needs to repeat the direction estimate. Therefore, it can bound the impact of directional analysis errors during user navigation. Borealis is also direction-guided user navigation.
In the process of direction analysis, it provides a good estimate of AP’s direction related to its physical location, meaning a user moves towards the AP and performs periodic direction estimates to tune its direction, and navigation ends when he reaches the AP.Figure3. Borealis Architecture Overview Figure3 is Borealis architecture overview. The main task of application layer is that the direction is displayed on the phone to guide the user navigation. This layer always allows the user to identify a target AP using its SSID, and directs the user to start the rotation. After analyzing the measured signal profile provided by orientation data (Compass Sensor) and WiFi RSS reports, this layer computes the AP direction and the confidence value of the current estimation.
In OS layer the deployment is we modified WiFi driver, we only scan the interested channel and accelerate the process by 10 seconds per rotation. It also saves the power that WiFi’s energy consumption is 14 times less. 5. Evaluation In this section, we implement many experiments in different kinds of outdoor environments and body shapes with natural gestures to verify Borealis’ effect on accuracy of estimation, confidence prediction, comparison to MinR, Offline and GUIDE analysis, and efficiency of indoor APs, Navigation and energy consumption. From all the experiments, we may safely come to the esults that Borealis is fairly accurate in simple LOS environment, but in complex LOS and NLOS it produces large errors in direction estimation. The cause is that the larger the angular error, the more likely that it is caused by multiple dips in the measured signal profile.
There is also a trend that the angular error scales inversely with the confidence value. For measuring AP direction, there are other three systems. MinR is the baseline algorithm for our proposed directional analysis. The angular error is 2 times the Borealis’ estimation error.Offline Analysis has the advantage of optimizing the direction estimation so that it can recognize certain patterns in complex environments and produce a more accurate decision. On the opposite aspect, it will cost a lot to realize.
So there is no doubt Borealis is a practical and effective solution for determining AP direction in real time. GUIDE is online and requires a very small set of measurements, but comparing with Borealis, the received signal strength of GUIDE degrades with the distance between the transmitter and receiver. But this does not always happen in practice.So Borealis definitely outperforms GUIDE.
As we know, the indoor scenario includes multipath propagation that degrade the accuracy of direction estimation. But the result shows that Borealis is effective and more accurate for locating indoor APs using outdoor signal measurements. The navigation overhead is the extra distance traveled to locate the AP. It’s computed by GPS records of navigation, and the shortest distance form GPS coordinates of the AP and the starting point. Since there are buildings and trailers affect the accuracy of Borealis’ direction estimates, the navigation paths generally ollow feasible walk paths efficiently.
The multiple propagation helps the user to identify a feasible path to the AP. Because Borealis’ signal measurements are passive during user rotation and do not involve packet transmission, therefore normal usage of Borealis will not influence the battery life of a smartphone. 6. Conclusion AP location is an important function in WiFi networks.
It can be used in hotspot management, access point detection, network security and so on. Borealis is an accurate and efficient system for locating WiFi access points in real time based on smartphone.It has the advantage that it can use off-the-shelf smartphones for locating accurately without the need of extensive measurements or expensive hardware. Not only efficiently used in outdoor environments, but effectively in locating indoor APs with low battery consumption.
Maybe it will address the network management issues in the future. Citation 1. Zengbin Zhang, Xia Zhou, Weile Zhang, Yuanyang Zhang, Gang Wang, Ben Y. Zhao, Haitao Zheng. I Am the Antenna: Accurate Outdoor AP Location using Smartphones. 2.
http://techtransfer. universityofcalifornia. edu/NCD/21959. html
