Kismet Wireless

Kismet Forums


Posted by:phoenix
Subject:tweaking spectools
Date:00:56:16 03/11/2010

> >
> > Hi All,
> >
> Since you decided to send this both here and email, I'm only going to bother replying once, here.
> > I stumbled across the spectools code, and found it very useful. Thanks a ton for this! I am using a 2.4x wi-spy device. I had a few questions about the same, and I was wondering if you could help me out:
> >
> > 1. Is it possible to tweak the frequency range? For example, if I only want to monitor a particular center frequency, how do I go about doing it? (i.e., all the samples from wispy would then corresponding to this frequency)
> There is code in the 24x specific file for setting it. There is currently no interface in the UI for controlling this.
> The 24x has somewhat limited range setting capabilities if I recall.

You are right. I was looking into the function wispy24x_usb_setposition(..) in wispy_hw_24x.c for this. This seems to be the relevant code ..

/* This is totally broken for anything but the default */
if (in_profile == 0) {
start_khz = WISPY24x_USB_DEF_H_MINKHZ;
res_hz = WISPY24x_USB_DEF_H_RESHZ;
best_s_m = WISPY24x_USB_DEF_STEP_MANT;
best_s_e = WISPY24x_USB_DEF_STEP_EXP;
} else {
/* Brute-force search the exponents and use the reverse of the function as the
* fitness test. Pick the resolution which is closest to our requested one. */
best_d = INT_MAX;
for (e = 0; e <= 4; e++) {
temp_m = WISPY24x_USB_RES_M_FROM_HZ(res_hz, e) & 0xFF;
temp_d = abs(res_hz - WISPY24x_USB_RES_HZ_FROM_ME(temp_m, e));

if (temp_d < best_d) {
best_d = temp_d;
best_s_m = temp_m;
best_s_e = e;


I am asking this question, based off the comment "this is totally broken for anything but default" --- can you let me know what needs to be fixed if I want to give any other parameters for start_khz and res_hz??

Also, could you please explain me a bit of what is going in this code i.e., what are best_d, best_s_m and best_s_e used for? Is there any restriction on the actual values of start_khz and res_hz ?

Also, could you explain, why the the number of samples fixed to 255? shouldn't this be calculated using (start-end)/res_hz ?

> >
> > 2. What is the current sampling rate? (i.e., how many RSSI values/sec can I get with the current code)? How much time does it take to move from one frequency to the other?
> Depends on the resolution of the scan, etc. Timing data is encoded in the sample records if you want to add instrumentation. There is currently no display for this.

Thanks for the info .. I will look into this.

> -m

Reply to this message