Smoothing filter for irregular update intervals
A technique for determining a received signal strength from multiple messages filters noise from the received signal to provide an accurate signal strength value. Advantageously, the more accurate output signal strength value can be used to identify movement of a station as well as estimate locations and direction of movement.
1. A system implemented in at least one of a memory or a processing device, comprising:
a dynamically windowed filter configured to be coupled to a radio receiving a current sample signal associated with a current sample signal strength value;
the dynamically windowed filter is configured to calculate a weighted input signal strength value from the current sample signal strength value;
the dynamically windowed filter is configured to calculate a decayed previous filter output value from a previous output value;
the dynamically windowed filter is configured to add the decayed previous filter output value to the weighted input signal strength value to define a filter output.
2. The system of claim 1 , wherein the previous output value is based on a a previous sample signal received by the radio, the previous sample signal associated with a previous sample signal strength value.
3. The system of claim 1 , wherein the dynamically windowed filter is configured to obtain a dynamic time window value, the dynamic time window being dynamically allocated as a factor of a weight applied to an input signal and a factor of a weight applied to a previous output signal.
4. The system of claim 1 , wherein the dynamically windowed filter is configured to calculate the decayed previous filter output value by multiplying the previous filter output value by e −t/τ , wherein t is a time elapsed value of a difference in time between receipt of a previous sample signal and receipt of the current sample signal and τ represents a dynamic time window value.
5. The system of claim 1 , wherein the dynamically windowed filter is configured to calculate the weighted input signal strength value by 1−e− t/τ , wherein t is a time elapsed value of a difference in time between receipt of a previous sample signal and receipt of the current sample signal and τ represents a dynamic time window value.
6. The system of claim 1 , wherein the dynamically windowed filter is configured to calculate the weighted input signal strength by calculating a function expressed as a Taylor polynomial approximation.
7. The system of claim 1 , wherein the filter output is a first filter output, the dynamically windowed filter configured to define a second filter output based on a subsequent sample signal strength value and the current output signal strength value, the subsequent sample signal strength value based on a suqbsequent sample signal received by the radio.
8. The system of claim 1 , wherein the dynamically windowed filter is configured to be coupled to a processor determining an elapsed time value by comparing (1) a first time value received from a clock when a previous sample signal is received with (2) a second time value received from the clock when the current sample signal was received.
9. A method, comprising:
receiving a previous output signal strength value, a current signal strength value, and a time elapsed value;
decaying the previous output signal strength value with (1) a first function of the time elapsed value and (2) a dynamic time window, to obtain a decayed previous output signal strength value;
weighting the current signal strength value with (1) a second function of the time elapsed value and (2) the dynamic time window, to obtain a weighted current signal strength value;
adding the decayed previous output signal strength value to the weighted current signal strength value to find a current output signal strength value.
10. The method of claim 9 , wherein the first function and the second function include a ratio of the time elapsed value to a dynamically allocated time constant value.
11. The method of claim 9 , wherein the first function is 1−e −t/τ , wherein t is the time elapsed value and τ represents a dynamic time window value.
12. The method of claim 9 , wherein the time elapsed value is determined as a difference between a first time when the current sample signal strength value is recorded and a second time when a previous sample signal strength value is recorded.
13. The method of claim 9 , further comprising:
receiving a subsequent signal strength value; and
adding a weighted subsequent signal strength value to a decayed current output signal strength value to find a subsequent output signal strength value.
14. A non-transitory processor-readable medium storing code representing instructions to be executed by a processor, the code comprising code to cause a processor to:
define a weighted input signal strength value based on a current sample signal strength value;
define a decayed previous filter output value based on a previous filter output value;
combine the decayed previous filter output value and the weighted input signal strength value to obtain a filter output.
15. The non-transitory processor-readable medium of claim 14 , wherein the previous filter output value is based on a previous sample signal, the previous sample signal associated with a previous sample signal strength value.
16. The non-transitory processor-readable medium of claim 14 , the code further comprising code to cause a processor to receive a dynamic time window value, the dynamic time window is dynamically allocated as a factor of a weight applied to the sample signal strength value and a factor of a weight applied to a previous output filter value.
17. The non-transitory processor-readable medium of claim 14 , wherein the decayed previous filter output value is defined by multiplying the previous filter output value by e −t/τ , wherein t is a time elapsed value of a difference in time between receipt of (1) a current sample signal associated with a current sample signal strength value and (2) a previous sample signal associated with previous filter output value, and τ represents a dynamic time window value.
18. The non-transitory processor-readable medium of claim 14 , wherein the weighted input signal strength value is defined by 1−e− t/τ , wherein t is a time elapsed value of a difference in time between receipt of a (1) a current sample signal associated with a current sample signal strength value and (2) a previous sample signal associated with previous filter output value and τ represents a dynamic time window value.
19. The non-transitory processor-readable medium of claim 14 , wherein the weighted input signal strength value is defined by calculating a function expressed as a Taylor polynomial approximation and based on the sample signal strength value.
20. An apparatus, comprising:
a dynamic window filter configured to receive (1) a current signal strength value, (2) a previous output signal strength value, (3) a dynamic time window value, and (4) a time elapsed value,
the dynamic window filter configured to define a decayed previous output signal strength value based on the previous output signal strength, a first function of the time elapsed value, and the dynamic time window value,
the dynamic window filter configured to define a weighted current signal strength value based on the current signal strength value, a second function of the time elapsed value, and the dynamic time window value,
the dynamic window filter configured to define a current output signal strength value based on the decayed previous output signal strength value and the weighted current signal strength value.