LDIL-MATLAB

A low-cost device-free indoor localization tool implemented in MATLAB, mostly based on LiFS
We aim to take better care of elders at home without expensive and complicated devices

Dependencies

linux-80211n-csitool
Hostapd, see our wiki - Hostapd Installation Guide
MATLAB toolbox
- signal processing toolbox
- global optimization toolbox
- optimization toolbox
- Having problems with MATLAB? See the wiki - Background Knowledge - MATLAB

In main.m, you need to change parameters according to the following scenarios:

.dat: select .dat file collected by Intel 5300 NIC
baseIndex,baseIndex1: the base packet index for csi_trace
Ci, Cj: location of AP and MP
noise:
- measurement noise
- Actually it should be optimized by genetic algorithm
At:
- When a target is located exactly on the LoS path, a link suffers large extra signal attenuation absorbed by the target
- Actually it should be optimized by genetic algorithm
ht: the distance from the highest point of the target to the wireless link
J: We couldn't figure out what J is in the essay, therefore we just set this to 1
genetic algorithm opts

All of the parameters above, and
y: collect all CSIeff and targetLocation
targetLocation: store all targetLocation from rough_location_estimate(f0, f, delta, F, O) as an array

>> main

Gradient descent algorithm
- In the essay, it use ga and then gd to optimize the target locaiton
- However, in this inplementation, we only use ga.
Multi-AP & multi-MP scenario
- Some part of code need to be modified to support multi-AP&MP scenario