**3. Positioning with smartphones: outdoor and indoor scenarios**

When outdoor scenarios are considered, smartphone technology can provide positions with a quite good level of accuracy, using the assisted GPS (A-GPS) system. Despite that, it is possible that the received GPS/GNSS signal is too noisy or not available at all, for example, if the user is in urban canyons or inside buildings: in these cases GNSS positioning is not possible.

Starting from that, many researchers have been investigating alternative solutions that consider different sensors (such as INS and images) and other technologies (e.g., Wi‐Fi, pedestrian tracking system, Bluetooth) in order to improve position accuracy and availability. A brief overview about accuracies obtainable today with a generic smartphone (chosen as representative) is made in the following subsections.

#### **3.1. Outdoor scenarios**

**CMOS imaging sensor characteristics Smartphone Sensor name Size** 

168 Smartphones from an Applied Research Perspective

Sony Exmor RS IMX378

Sony Exmor RS IMX315

Sony Exmor RS IMX\*

Sony Exmor RS IMX\*

Sony Exmor RS IMX240

Sony Exmor RS IMX260

Samsung Isocell S5K2L1

RS IMX286

Sony Exmor RS IMX298

Sony Exmor RS IMX300

Sony Exmor RS IMX400

RS IMX234

**Table 1.** CMOS image sensor characteristics for commercial camera phones.

**LG G4 e G5** Sony Exmor

**Huawei P9** Sony Exmor

**Google Pixel/ BlackBerry Key One**

**Apple iPhone 6S**

**Apple iPhone 7**

**Apple iPhone 7 Plus**

**Samsung Galaxy S6, S6 Edge(+)**

**Samsung Galaxy S7, S7 Edge**

**OnePlus 3T/LGV20/ Huawei Mate 8/Asus Zenfone 3**

**Sony Xperia XZ**

**Sony Xperia XZ Premium (coming soon)**

**(diagonal) [mm]**

**dpix [µm] CMOS** 

**technology**

7.81 1.55 BSI CMOS 6.25 **×** 4.69 4032 **×** 3024 12.2

6.15 1.22 BSI CMOS 4.92 **×** 3.70 4032 **×** 3024 12.2

6.15 n/a n/a n/a n/a 12

5 n/a n/a n/a n/a 12

6.83 1.12 BSI CMOS 5.95 **×** 3.35 5312 **×** 2988 15.9

7.06 1.4 BSI CMOS 5.64 **×** 4.23 4032 **×** 3024 12.2

7.06 1.4 ISOCELL 5.64 **×** 4.23 4033 **×** 3024 12.2

6.2 1.25 BSI CMOS 4.96 **×** 3.72 3968 **×** 2976 11.8

6.4 1.12 BSI CMOS 5.16 **×** 3.87 4608 **×** 3456 15.9

7.87 1.08 BSI CMOS 6.46 **×** 4.47 5984 **×** 4140 24.8

7.73 1.22 BSI CMOS 6.17 **×** 4.63 5056 **×** 3792 19.2

6.83 1.12 BSI CMOS 5.95 **×** 3.35 5312 **×** 2988 15.9

5.96 **×** 4.47 5520 **×** 4140

6.46 **×** 3.64 5984 **×** 3366

(4:3 mode)

(16:9 mode)

22.8

20.1

**Sensor dimensions [mm × mm]**

**Image dimension [pix × pix]**

**MP**

## *3.1.1. GPS/GNSS only*

As said in Section 2.1, starting from the end of 2016, it is possible to acquire raw GNSS measurements from smartphones: the main problem is that only Android Nougat OS allows to extract these information. Thus, in this section the attention is focused only on internal solutions provided by software installed on smartphones. In order to analyze the precision obtainable today with GNSS internal chipset, some tests were performed.

The tests took place in the same places described in [11] considering two different scenarios: an open outdoor area to represent "ideal" conditions (**Figure 2**, left) and another area (one of the courts in Politecnico di Torino campus) with characteristics of urban canyon (**Figure 2**, right). The line in **Figure 2** ‐ right shows a particular track where it is possible to find an area with a limited satellite visibility (similar to urban canyon conditions) and many windows that create multipath due to their high reflectivity.

Dynamic tests were performed in these areas by walking along the same path with the smartphones mounted on a special "two-hands" support as shown in **Figure 3**. The entire data collection system includes:


GNSS data positions were recorded during the surveys considering a one-second sample rate, using a dedicated app that stores the National Marine Electronics Association (NMEA)

**Figure 2.** Test site and track: an open sky area (left) and an urban canyon (right).

**Figure 3.** The two-hands support system developed at Politecnico di Torino.

GGA messages in an ASCII file. All results were compared with a "ground‐truth" obtained through the continuous tracking of the smartphone position with a total station, thanks to the retro‐reflector installed on the "two‐hands" support. In this way, a millimeter accuracy was obtained, considering and estimating the level‐arm offset between the instruments.

The NMEA sentences were analyzed and compared with the reference trajectories using software written in MATLAB.

The horizontal positioning errors of the representative receiver are shown in **Figure 4** for the urban canyon environment.

In order to have a more complete analysis from a statistical point of view, the most significant statistical parameters have been summarized in **Table 2** for the urban canyon and open area test locations.

As expected, then, it is generally possible to affirm that some environmental characteristics, such as obstacles, multipath effects coupled with the number of trackable satellites, play a crucial role in the accuracy determination of the smartphone positioning.

Positioning Techniques with Smartphone Technology: Performances and Methodologies... http://dx.doi.org/10.5772/intechopen.69679 171

**Figure 4.** 2D performances of internal GPS sensor.

GGA messages in an ASCII file. All results were compared with a "ground‐truth" obtained through the continuous tracking of the smartphone position with a total station, thanks to the retro‐reflector installed on the "two‐hands" support. In this way, a millimeter accuracy was

The NMEA sentences were analyzed and compared with the reference trajectories using soft-

The horizontal positioning errors of the representative receiver are shown in **Figure 4** for the

In order to have a more complete analysis from a statistical point of view, the most significant statistical parameters have been summarized in **Table 2** for the urban canyon and open area

As expected, then, it is generally possible to affirm that some environmental characteristics, such as obstacles, multipath effects coupled with the number of trackable satellites, play a

crucial role in the accuracy determination of the smartphone positioning.

obtained, considering and estimating the level‐arm offset between the instruments.

**Figure 2.** Test site and track: an open sky area (left) and an urban canyon (right).

170 Smartphones from an Applied Research Perspective

**Figure 3.** The two-hands support system developed at Politecnico di Torino.

ware written in MATLAB.

urban canyon environment.

test locations.


**Table 2.** Error statistics in urban canyon and open area environments.

However, precision and accuracy improvements could be increased by computing a differential positioning solution, considering the raw measurements obtainable from internal sensors.
