I am using NRF24L01 and ZigBee modules for wireless communication. I kept the modules in the tapered closed aluminum cylinder for sensor data transmission. The signal was being received when the module was placed at some specific positions (closed to the center). It stopped/slowed down the transmission when it was at other locations. Does this mean it is working as a waveguide?
It at least means it's working as a cavity.
There are three easy to describe regimes that a closed metal box may work in, when there's an RF field inside (and a few more that aren't so simple).
a) Waveguide, or single mode
The two small dimensions of a tube are comparable to half the wavelength. EM waves will propagate with low loss along the long dimension. Commercial waveguides are designed with sizes to be used for specific frequencies so that only one mode propagates. This is important, as different modes travel at different speeds. (You can use multi-mode waveguide, but it's a specialist application and tricky to use.)
b) Large cavity, or multimode
At least two box dimensions are larger than half a wavelength. Many modes are possible, and you can't easily predict which mode most energy is going to be in, especially if there are things in the box disturbing the field. This is the microwave oven regime, where food hot-spots and cold-spots are more or less unpredictable with food position.
Note that industrial RF heating usually uses a single mode cavity, where the heating occurs in a predictable position.
c) Small cavity, or evanescent
The two small dimensions are less than a small fraction of a wavelength. Energy does not propagate as EM fields, but can couple as electric fields with strong attenuation. This is the regime used for the shield round a high gain IF amplifier for instance, where the isolation from the high gain output back to the input must be strongly attenuated to avoid it picking up its own output.
It sounds like you're in the multimode regime, where detailed changes to positions move you in and out of signal nulls.