Please guide me about the difference between SINR threshold and Minimum discernable signal.
From my search I have come across the following.
1) A signal can be decoded if the SINR of the received signal is higher than the SINR threshold. Does it mean that we should not be concerned about the minimum required power, and that if the received signal satisfies the SINR threshold, it will be successfully decoded?
2) I also have come across the idea of minimum discernable level. For instance -70 db is considered acceptable for some types of communication.
Which of the two I should follow. As in the first case, I get very low transmit powers and still satisfy the SINR threshold, while the transmit powers in the 2nd case are way too high compared to the first case.
The minimum discernible/detectable signal level is the minimum signal strength at which the receiver hardware can notice that there exist anything else than noise. But just because one can detect the presence of a signal, it doesn’t mean that one can decode the information that is encoded into the signal.
Depending on what modulation and coding scheme that is utilized, there will be an SINR threshold at which the decoding works (most of the time, there is always a non-zero probability of error). One basically selects the acceptable error probability and compute what SINR level is needed to achieve it.
I think you mostly have to care about the SINR threshold.
Dear Gul Lakhta ,
Adding to Emil Björnson , the threshold value of the SINR sets the level of the limit of the highest allowed bit error rate. For all types of communication systems, the bit error rate depends on the signal to noise ratio. As this ratio increases the bit error rate decreases. There is no communication system with absolute zero bit error rate but there is acceptable quality of service which sets the upper limit of the bit error rate. It is normally recommended that for voice communication the acceptable bit error rate must be no more than 1x10^-3 and for data the acceptable error rate.is about 1x10^-6.
The noise is generated normally in the receiver which receives the signal power.
If there is no noise the detector would correctly detect the received symbols.
Due to noise superposition on the received symbol signals there will be some probability that the detector will falsely detect the symbol. It is so that the noise superimposed on the symbol signal will cause to shift the symbol in the decision regions of the adjacent symbols.
In order to refer the importance of the signal to noise ration for the symbol detection process you have to refer to the so called constellation diagrams in which every symbol appears appears inside noise contaminated area. The noise appears as light flake around the central position of the symbol.
Best wishes
Thank you Emil Björnson Abdelhalim abdelnaby Zekry for your valuable comments.
Abdelhalim abdelnaby Zekry: Sir does "If there is no noise the detector would correctly detect the received symbols" mean that if there is no noise, the receiver can still decode a signal even if the transmit power of the transmitter is set based only on the Signal to Interference ratio while ignoring the minimum discernable level?
Dear Gul Lakhta ,
There is something basic which you you would like to consider. From the point view of the signals, the interference signal acts in a similar way to noise. When it is added to the symbol signal it will cause that it may be falsely detected if its value is increases such that the combined signal will be located in the neighboring decision regions of the adjacent symbols in the constellation diagram. So the interference and noise have the same effect in the symbol error.
Therefore one speaks not only from S/N ratio but also from the S/(N+I).
The noise has Gaussian amplitude distribution while the interference may not have such distribution because of the nature of the interference.
The interference signal sources are normally known and then they can be subtracted from the composite signal. If after interference cancellation there will be residual interference it will be treated as a noise concerning its effect on the symbol error. In the sense that it will set the S/I +N minimum level to achieve specific bit error rate. You see if N=0 then S/I will set this minimum.
Hope I could answer satisfactorily you question.
Best wishes
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity