Co-site interference - Simulation Approach.
时间：04-04
整理：3721RD
点击：
Hello all,
First of all, I apologize for evey mistake or blunder I make within this post but honestly its my very first time here
Problem that I experience it kind of hard to explain but I will try my best. My employer wants to make a simulation when few transmitters with the same waveform (modulation, coding scheme etc.) operates in given frequency range. Every transmitter has different center frequency. There is one receiver which creates a link with one of the transmitter (they are operating at the same frequency). The goal is to obtain an information, with what maximum power, other transmitter working in adjactent channel, could work that SNR in link between two mentioned devices doesn't change. The parametr's we know are:
- transmitter mask,
- receiver selectivity mask,
- channel bandwidth = 1.25 MHz.
As an input data to my algoritm I used measurements which are made using some radio devices (doesn't matter which in my opinion). Also I used the tx and rx masks.
The algorithms goes as follows:
- there is a database which stores information about noise power and IQ data (if exists) in each channel (data from tx masks),
- by adding new transmitter power in each channel, values are beeing summed togateher,
- receiver takes data from database and using rx mask introduce attnuation in each channel (despite of main channel),
- AGC simulation (detector in the receiver checks how much power is in the bandwith and adjust the gain, worth to metion that it is wideband square-law detector),
- after detector, signal from main channel is filtered out and signal reaches ADC,
- ADC checks if signal is full scaled, if not SNR is decreased by the difference between optimal ADC input power and actual signal power.
Problem is that when I run my simulation and compare measurements from the hardware, results diffrence reaches apx. 20 dB.
Question is, what other details should I consider in my simulation? What is the problem? I will be glad to hear professionals comments under this post.
P.S. I attach image with the plot of my results (simulated) and measured curve.
Thanks in the advance,
Martin
First of all, I apologize for evey mistake or blunder I make within this post but honestly its my very first time here
Problem that I experience it kind of hard to explain but I will try my best. My employer wants to make a simulation when few transmitters with the same waveform (modulation, coding scheme etc.) operates in given frequency range. Every transmitter has different center frequency. There is one receiver which creates a link with one of the transmitter (they are operating at the same frequency). The goal is to obtain an information, with what maximum power, other transmitter working in adjactent channel, could work that SNR in link between two mentioned devices doesn't change. The parametr's we know are:
- transmitter mask,
- receiver selectivity mask,
- channel bandwidth = 1.25 MHz.
As an input data to my algoritm I used measurements which are made using some radio devices (doesn't matter which in my opinion). Also I used the tx and rx masks.
The algorithms goes as follows:
- there is a database which stores information about noise power and IQ data (if exists) in each channel (data from tx masks),
- by adding new transmitter power in each channel, values are beeing summed togateher,
- receiver takes data from database and using rx mask introduce attnuation in each channel (despite of main channel),
- AGC simulation (detector in the receiver checks how much power is in the bandwith and adjust the gain, worth to metion that it is wideband square-law detector),
- after detector, signal from main channel is filtered out and signal reaches ADC,
- ADC checks if signal is full scaled, if not SNR is decreased by the difference between optimal ADC input power and actual signal power.
Problem is that when I run my simulation and compare measurements from the hardware, results diffrence reaches apx. 20 dB.
Question is, what other details should I consider in my simulation? What is the problem? I will be glad to hear professionals comments under this post.
P.S. I attach image with the plot of my results (simulated) and measured curve.
Thanks in the advance,
Martin
