Understanding the Concept:
A matched filter is an optimal linear filter designed to maximize the output signal-to-noise ratio (SNR) at a specific sampling moment in the presence of white Gaussian noise. The impulse response ($h(t)$) of a matched filter is a time-reversed and delayed version of the known clean input signal ($s(t)$):
\[
h(t) = s(T - t)
\]
where $T$ is the total symbol duration.
Step 1: Finding the Filter Output using Convolution
The output $y(t)$ of any linear time-invariant (LTI) system is found by convolving its input signal $x(t)$ with its impulse response $h(t)$:
\[
y(t) = x(t) * h(t) = \int_{-\infty}^{\infty} x(\tau) h(t - \tau) \, d\tau
\]
Now substitute the matched filter's characteristic impulse response, $h(t) = s(T - t)$, into this convolution equation:
\[
y(t) = \int_{-\infty}^{\infty} x(\tau) s(T - (t - \tau)) \, d\tau = \int_{-\infty}^{\infty} x(\tau) s(T - t + \tau) \, d\tau
\]
Step 2: Evaluating the Output at the Optimal Sampling Instant
We evaluate the filter's output at the optimal sampling moment, $t = T$:
\[
y(T) = \int_{0}^{T} x(\tau) s(T - T + \tau) \, d\tau = \int_{0}^{T} x(\tau) s(\tau) \, d\tau
\]
Step 3: Comparing with a Correlator Receiver
Let us look at how a standard correlator receiver operates:
• It multiplies the incoming noisy signal $x(\tau)$ directly by a locally generated clean copy of the signal $s(\tau)$.
• It integrates this product over the symbol time window from $0$ to $T$.
The mathematical expression for this correlator operation is:
\[
\text{Output}_{\text{correlator}} = \int_{0}^{T} x(\tau) s(\tau) \, d\tau
\]
Since the mathematical operations are identical, a matched filter sampled at $t = T$ is completely equivalent to a correlator receiver.
Step 4: Disproving alternative choices
• Envelope Detector: An asynchronous circuit built from a diode and capacitor used to extract the modulation envelope from AM signals, which does not maximize SNR using signal templates.
• PLL (Phase-Locked Loop): A closed-loop feedback control system used for tracking phase and frequency, not for optimal pulse detection.
• Equalizer: Used to reverse the distorting effects of a communication channel, rather than maximizing SNR for a known pulse shape in white noise.