Exponential Fourier Series

Introduction

The exponential (or complex) form of the Fourier series is the most compact, elegant, and mathematically powerful representation of periodic signals. While the trigonometric form (sines and cosines) might feel more intuitive initially, the exponential form using $e^{jn\omega_0 t}$ is preferred in advanced signal processing for several compelling reasons: it leads to simpler derivations, extends naturally to the Fourier transform, and provides a unified treatment of magnitude and phase through complex coefficients.

For B.Tech students, mastering the exponential Fourier series is essential because it is the form used in most textbooks for deriving the Fourier transform, understanding spectral analysis, and working with system frequency responses. Once you are comfortable with complex exponentials, the entire landscape of signal analysis becomes cleaner and more unified.

The Exponential Fourier Series Representation

Any periodic signal $x(t)$ with period $T_0$ and fundamental frequency $\omega_0 = 2\pi/T_0$ can be represented as:

$$x(t) = \sum_{n=-\infty}^{\infty} c_n e^{jn\omega_0 t}$$

where the complex Fourier coefficients $c_n$ are computed as:

$$c_n = \frac{1}{T_0}\int_{T_0} x(t) e^{-jn\omega_0 t} dt$$

The integration is over any complete period (e.g., from $0$ to $T_0$ or from $-T_0/2$ to $T_0/2$).

Understanding the Formula

The coefficient $c_n$ is found by correlating (taking the inner product of) the signal $x(t)$ with the complex exponential $e^{jn\omega_0 t}$. This correlation extracts the component of $x(t)$ at frequency $n\omega_0$. The factor $1/T_0$ normalizes by the period length.

Each $c_n$ is generally a complex number: $c_n = |c_n|e^{j\angle c_n}$

• $|c_n|$ gives the amplitude of the $n$th harmonic
• $\angle c_n$ gives the phase of the $n$th harmonic
• $n > 0$: positive frequency components
• $n < 0$: negative frequency components
• $n = 0$: DC component (average value)

Properties of Exponential Fourier Coefficients

Conjugate Symmetry for Real Signals

If $x(t)$ is real-valued, then:

$$c_{-n} = c_n^*$$

This means:

• $|c_{-n}| = |c_n|$ (magnitude spectrum is even)
• $\angle c_{-n} = -\angle c_n$ (phase spectrum is odd)
• $c_0$ is always real (DC value = average of the signal)

Parseval's Theorem

The average power of $x(t)$ equals the sum of powers of all harmonics:

$$P = \frac{1}{T_0}\int_{T_0}|x(t)|^2 dt = \sum_{n=-\infty}^{\infty}|c_n|^2$$

This is the power version of energy conservation — the total power is distributed among all frequency components.

DC Component

$$c_0 = \frac{1}{T_0}\int_{T_0} x(t) \, dt$$

This is simply the time-average (mean value) of the signal.

Relationship to Trigonometric Form

The exponential and trigonometric forms are directly related. If:

$$x(t) = \frac{a_0}{2} + \sum_{n=1}^{\infty}[a_n\cos(n\omega_0 t) + b_n\sin(n\omega_0 t)]$$

Then the exponential coefficients are:

$$c_0 = \frac{a_0}{2}, \quad c_n = \frac{a_n - jb_n}{2}, \quad c_{-n} = \frac{a_n + jb_n}{2}$$

Conversely: $a_n = 2\text{Re}(c_n) = c_n + c_{-n}$ and $b_n = -2\text{Im}(c_n) = j(c_n - c_{-n})$

The amplitude of the $n$th harmonic in the trigonometric form is:

$$A_n = \sqrt{a_n^2 + b_n^2} = 2|c_n|$$

Line Spectra (Spectral Representation)

The exponential Fourier series naturally leads to spectral plots:

Magnitude spectrum: Plot $|c_n|$ vs $n\omega_0$ (or $nf_0$). Shows the amplitude at each harmonic frequency.

Phase spectrum: Plot $\angle c_n$ vs $n\omega_0$. Shows the phase at each harmonic frequency.

Together, these form the line spectrum (or discrete spectrum) of a periodic signal. The term "line" refers to the fact that energy exists only at discrete frequencies $0, \pm f_0, \pm 2f_0, \pm 3f_0, \ldots$

Worked Examples

Example 1: Periodic Rectangular Pulse Train

Consider a pulse train with period $T_0$, pulse width $\tau$, and amplitude $A$:

$$c_n = \frac{1}{T_0}\int_{-\tau/2}^{\tau/2} A e^{-jn\omega_0 t} dt$$

$$= \frac{A}{T_0} \left[\frac{e^{-jn\omega_0 t}}{-jn\omega_0}\right]_{-\tau/2}^{\tau/2} = \frac{A}{T_0} \cdot \frac{2\sin(n\omega_0\tau/2)}{n\omega_0}$$

$$= \frac{A\tau}{T_0} \cdot \text{sinc}\left(\frac{n\tau}{T_0}\right)$$

where $\text{sinc}(x) = \sin(\pi x)/(\pi x)$. The magnitude spectrum has a sinc envelope with zeros at $n = T_0/\tau, 2T_0/\tau, \ldots$

This is one of the most important results in signals and systems — it shows that a periodic pulse train has a discrete spectrum whose envelope is the sinc function.

Example 2: Full-Wave Rectified Cosine

For $x(t) = |\cos(\omega_0 t)|$ (period $T_0/2$, fundamental $2\omega_0$):

$$c_n = \frac{2}{\pi} \cdot \frac{(-1)^{n+1}}{4n^2 - 1}$$

The coefficients decay as $1/n^2$ because the full-wave rectified cosine is continuous (smoother than a square wave).

Example 3: Sawtooth Wave

For a sawtooth wave of amplitude $A$ and period $T_0$:

$$c_n = \frac{jA}{2\pi n} \cdot (-1)^n = \frac{jA(-1)^n}{2\pi n}, \quad n \neq 0$$

$$c_0 = 0 \text{ (if symmetric about zero)}$$

The magnitude decays as $1/|n|$ (slow — the sawtooth has a jump discontinuity).

Effect of Signal Properties on Spectrum

From Fourier Series to Fourier Transform

The exponential Fourier series for periodic signals can be extended to aperiodic signals by letting $T_0 \to \infty$. As the period increases:

• Harmonic spacing $\omega_0 = 2\pi/T_0$ shrinks to zero
• Discrete frequencies become a continuous spectrum
• The sum becomes an integral
• $T_0 \cdot c_n$ becomes the Fourier transform $X(\omega)$

This limiting process gives: $$X(\omega) = \int_{-\infty}^{\infty} x(t)e^{-j\omega t}dt$$

This is the foundation for the transition from Fourier series (periodic signals) to the Fourier transform (aperiodic signals).

Key Takeaways

• The exponential Fourier series $x(t) = \sum c_n e^{jn\omega_0 t}$ is the most general and compact representation
• Coefficients: $c_n = \frac{1}{T_0}\int_{T_0} x(t)e^{-jn\omega_0 t}dt$ — correlation with complex exponentials
• For real signals: $c_{-n} = c_n^*$ (conjugate symmetry) — magnitude is even, phase is odd
• Parseval's theorem: average power = $\sum|c_n|^2$ — power is distributed among harmonics
• The line spectrum shows amplitude and phase at discrete harmonic frequencies
• Coefficient decay rate reflects signal smoothness: discontinuous → $1/n$, continuous → $1/n^2$, smooth → $1/n^k$
• The Fourier transform emerges naturally as $T_0 \to \infty$ in the exponential series