RF interference (RFI) between UHF personal radios and biomedical monitoring sensors

This work quantifies the RF interference (RFI) potential of medium-power UHF personal radios operating in close proximity to susceptible body-worn conductors. With the increasing awareness of safety hazards in the workplace, there is a need to provide continuous, remote biomedical monitoring by radio for personnel such as fire-fighters. Typically, the subject wears appropriate instrumentation interfaced to their medium-power personal radio (PR), or a separate and dedicated biomedical telemeter. This poses constraints on the instrumentation, whereby it must remain interference-free. RFI can originate in unwanted coupling between transmitting PR antennas and any proximate circuitry such as ECG lead sets and PCBs carrying sensitive amplifiers; coupled wires can also affect SAR distribution. The effects of coupling between short, vertical and tilted open-circuited wires laid over a muscle tissue slab (to mimic human ECG radio telemetric monitoring) and a nearby 435 MHz λ/2 dipole were investigated using a custom FDTD simulator written in 'C': COCA (Code for Coupled Antennas). The undesirable effects of analysing tilted elements using the Yee algorithm were overcome by applying Dey-Mittra special update equations. This approach was validated initially by comparing the input impedances and resonant frequencies of isolated dipoles oriented vertically and at 45° in the modelling space (a 1 mm Yee cell lattice]. The resonant-frequency shift was zero and the input resistance changed by only 1.72 % when applying the Dey-Mittra technique: see Fig. 1, where Remcom's XFDTD 5.3 simulator is used for comparison. An adult body torso, represented by a homogeneous muscle slab 70 cm tall × 30 cm wide × 15 cm deep, was introduced: Fig. 2. The effects of tissue / wire coupling on the dipole's input impedance was investigated by varying the horizontal separation between the dipole and wire, from 10-100 mm. The wire length was 188 mm, representative of that used in ECG radio-monitoring practice: Fig. 3. For the dipole and body only, minimum separation results in significant detuning, with an inductive input reactance and high body loss. Reactance falls with increasing separation, as does the feed resistance initially as absorptive loss reduces, minima being reached at 30 and 40 mm, respectively. Thereafter, the feed reactance remains inductive and the input resistance begins to rise towards the free-space value at about 120 mm spacing. Introducing the wire de-emphasises direct body coupling: close-in feed reactance is virtually unchanged from the isolated dipole case and direct absorption reduced, as evidenced by the smoother, vertically shifted (by about 2Ω) resistance plots. Note that the feed resistance profiles of the dipole adjacent to both wires are almost identical. The difference between the reactance plots is low, less than 0.5 Ω at separations greater than 40 mm. The peak currents induced in the monitor wire were calculated, for a dipole feed-point drive normalised to a 100 mA: close-in values of 22 mA and 7 mA were found for the vertical and tilted wire, respectively. The full Colloquium presentation will include current distributions found along the wire length, and the effect of close coupling on PR antenna polar patterns and body SAR.