To normalize a signal to zero mean and unit variance, you can follow these steps:

1. Compute the mean of the signal by summing up all the values in the signal and dividing by the total number of values in the signal.

2. Subtract the mean from each value in the signal. This will ensure that the signal has a zero mean.

3. Compute the standard deviation of the signal by taking the square root of the sum of the squared differences between each value and the mean, divided by the total number of values in the signal.

4. Divide each value in the signal by the standard deviation. This will ensure that the signal has a unit variance.

Alternatively, you can use a library like NumPy in Python to perform these operations automatically. Here's an example code snippet: 

import numpy as np

# Assume signal is stored in a variable called "signal"
normalized_signal = (signal - np.mean(signal)) / np.std(signal)

Note that normalization is a common preprocessing step for many machine learning algorithms, as it can help improve model performance and convergence.