adjust the level in a Spectrum Analyzer

I have the Following situation, I want to measure a signal +13 dBm of level at a Frequency of 1885 MHz , I want to test this Signal with a FSH3 Spectrum Analyzer from Rohde&Schwarz (this is a Handheld Spectrum Analyzer) when i adjust the Reference level to +20dBm I could measure the Signal as +12dBm (Regarding the Cable Loss) but the Apeadred noise around the Signal are at -50dBm (and this is not True) and when I tried to decrease the Ref. Level to 0dbm and decrease also the RBW, I found that the Measured Signal going to -4dBm and the noise around it is -70dBm ( and this also not true). I need to Know How can i Measure this Signal (+13dBm) according to the Real noise around it ( Which is around -90 to -100dBm or less)

These hand held spectrum analyzers are really not that good, as far as noise figure and dynamic range goes.

Firstly, you do realize that the noise floor measurement will change dramatically as you change the RBW. That is pure math. If you have noise that is at -100 dBm/Hz of bandwidth, and you have your RBW set at 1 KHz, you are letting in 1000 times (-100 dbm/Hz) of noise, or in other words, -70 dBm/KHz will be shown on the screen face. So, depending on the RBW setting you really do not need a huge dynamic range.

If you are looking at just random background noise, you can use a narrowband notch filter to filter out the carrier but pass the noise on either side. You measure the noise level, then crank in some attenuation and remove the notch filter to measure the carrier level.

ok, thank you but I want to ask another Question, why there is a decreasing on the Level of the Signal when i decrease the RBW? ( and it is a big different in level)

I USED THE SPECTRUM ANALYZER IN MY PROJECT,BUT THEIR WERE PROB'S CMING RELATED 2 POWER MEASURING,
SO V FOUND OUT DA OFFSET POWER LEVEL AND ADDED IN RESULTS AND GOT CORRECT RESLTS.

U CAN USE POWER METERS FOR POWER
MEASUREMENT OR VECTOR NETWRK ANALYZER FOR THIS................

If it is a modulated signal, then by decreasing the RBW there is less actual power getting into the spectrum analyzer's IF power detector. Imagine a spread spectrum signal...obviously if you use a lower RBW, less of the spread signal will get through instantaneously.

If you are talking about an unmodulated CW signal, changing the RBW should not change the power level. However, sometimes if things are compressed, or if signal is not too stable (bouncing around a little in frequency) a RBW that is too small may not be able to "catch it".