Abstract
This paper proposes to process the tracking and demodulation of two wireless signals at the same time in a single channel, via a BandPass Sampling front-end and any antenna arrangement. We propose two new multi-signal-PLL designs that can achieve this. These PLLs can track these two signals without losing lock (or cycle slipping) to any one of them, irrespective of the frequency step/gap between the two input signals frequencies (pull-in range). The Bluetooth (BT) and the GPS-L1CA (GPS) signals are chosen to test our two PLL designs. This is because these two signals are diverse, challenging and used extensively in Smartphones. The design challenge was to continually lock to the GPS signal while BT is communicating using the same receive chain, thus saving power and silicon. This has necessitated the design of a PLL with wide pull-in range and fast switching cycle. Our new multi-mode PLL’s design is based on integrating an adaptive Frequency Estimator (FE) into a standard version of the Costas PLL. The FE provides an estimated frequency of the received signal to modify the free-running frequency of the NCO that will enhance the performance of the PLL by increasing its lock-in range. To estimate the frequency, in our first PLL design approach, we used an adaptive IIR notch filter model using the least square method to (1) solve the minimization problem that relates the current received signal with the previous output of the filter, and (2) update the filter parameter that estimate the fundamental frequency. For our second PLL design approach, a numerical differentiation model is used to perform a fourth-order differentiation on the received signal samples in order to calculate the fundamental frequency. Our two Adaptive Multi-Mode PLLs (AMM-PLL) designs have been MATLAB simulated against various test scenarios and can: 1. track multi-signal with up to 10 KHz frequency steps (hold-in range). 2. perform accurate switching between different operation-modes (eg. BT and GPS). 3. have a stable phase-error that reaches 0.05 rad and 0.22 rad based on using IIR notch filter and numerical differentiation models respectively. 4. have a transition time for switching between the two AMM-PLL modes of 38 µsec and 25 µsec in the IIR notch filter and numerical differentiation models respectively. 5. Produce zero BER of the demodulated data of both signals at SNR greater than 12 dB. Our proposed solution, when implemented in Smartphones sensors, will result in considerable power saving, and reduce cost and size of the overall solution.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have