Asymmetric Interference Perception Method in Short-Range Wireless Systems

In the 2.4 GHz band, the channel of a short-range link is usually selected at one end; in Bluetooth Low Energy, a peripheral’s channel classification is advisory. This paper examines, for one instrumented case, whether the two ends of a link perceive interference similarly. A purpose-built protocol on two nRF52840 devices reported each end’s noise floor once per adaptation window. On a 4.2 m indoor link with one swept narrowband interferer, 656 time-paired windows were obtained. In the 614 quiet windows, the transmitter read a standing +6.30 dB more than the receiver, whereas a co-location control gave a mean offset of 0.66 dB. In the 42 windows within fifteen interference episodes, the difference had a standard deviation of 19.2 dB, and the two ends classified the channel differently in 38. Co-located devices also differed by a median of 9 dB under a directed source, so these differences cannot be attributed wholly to position. With two simultaneously sampled receive chains 60 cm apart, the levels received from a moving source were uncorrelated. A negotiation in which the receiver selects among the transmitter’s candidate channels completed 17 of 17 changes on a quiet bench and 43 of 44 under interference; in all 43, however, it selected the channel that transmitter-only selection would have used. Endpoint perception therefore differed substantially in this configuration, whereas an advantage of the method over simpler alternatives was not demonstrated.