The Network Theory

A circuit is a graph. Nodes are junctions. Edges are components. The laws of circuit theory are the laws of graph theory applied to physical networks. This realization—simple, profound, almost obvious in hindsight—unifies everything.

Graphs and Circuits

Every circuit maps naturally to a graph. Each junction of three or more wires is a node. Each component—resistor, capacitor, source—is an edge connecting two nodes. Directed edges carry the convention of current flow. Undirected edges suffice for analysis where direction is irrelevant.

The graph tells you everything you need to know about a circuit's structure before you write a single equation. The number of nodes determines how many independent junction equations you need. The number of edges and the topology determine how many independent loop equations exist. Euler's formula for planar graphs—$V - E + F = 1$ (ignoring the external face)—gives you the mesh count directly. Topology precedes analysis.

This is not a metaphor. A resistor network and a network of water pipes, a network of roads, and a neural network share the same mathematical structure because they are all graphs. The equations are identical. The solution methods are identical. What changes is the physical interpretation of the edge weights.

Nodal Analysis

Nodal analysis is the graph-theoretic approach par excellence. You assign a voltage variable to every non-reference node. You write KCL at each node: the sum of currents leaving through every connected edge equals zero. Each current is expressed through the branch's constitutive relation—Ohm's law for resistors, $I = C\frac{dV}{dt}$ for capacitors, $I = \frac{1}{L}\int V,dt$ for inductors.

The result is a system of linear equations. In matrix form: $G\mathbf{V} = \mathbf{I}$, where $G$ is the conductance matrix (symmetric, positive-definite for passive circuits), $\mathbf{V}$ is the vector of node voltages, and $\mathbf{I}$ is the vector of injected currents. This matrix exists because the graph tells you how nodes connect, and the component values tell you how strongly they couple.

Solving this system gives you every node voltage. From there, every branch current follows. One graph, one matrix, one solution.

Mesh Analysis

Mesh analysis takes the dual perspective. Instead of nodes, you define loop currents for every independent mesh (window) in a planar graph. KVL around each mesh gives equations where the self-resistance of a mesh appears on the diagonal and mutual resistances between adjacent meshes appear off-diagonal.

The two approaches are duals of each other—node and mesh, voltage and current, KCL and KVL. In graph theory terms, nodal analysis works on the primal graph and mesh analysis works on the dual graph. They are the same problem viewed from opposite sides of the surface.

Beyond Planar Graphs

Not all circuits are planar. Some cannot be drawn on a plane without crossing edges. Karnaugh's theorem tells us which graphs are planar. For non-planar circuits, mesh analysis in its simple form fails, but nodal analysis still works. The graph-theoretic perspective clarifies this: nodal analysis depends only on the node-edge incidence matrix, which exists for any graph. Mesh analysis depends on a cycle basis, which requires more care in the non-planar case.

Network Theorems as Graph Identities

Thevenin's theorem, Norton's theorem, superposition—these are not independent results. They are consequences of linearity and the graph structure. Thevenin's theorem says: any linear network, viewed from any two nodes, is equivalent to a voltage source in series with a resistor. This is true because the node-voltage equations are linear, and linear systems have equivalent representations.

Superposition says: the response to multiple sources equals the sum of responses to each source individually. This is true because the conductance matrix $G$ is linear. $G\mathbf{V} = \mathbf{I}_1 + \mathbf{I}_2$ implies $\mathbf{V} = G^{-1}(\mathbf{I}_1 + \mathbf{I}_2) = G^{-1}\mathbf{I}_1 + G^{-1}\mathbf{I}_2$. The inverse of a matrix is linear. That is the entire proof.

What Network Theory Gains

Seeing circuits as graphs gives you:

• A clear separation between topology and component values
• Insight into which analyses are possible (planar vs. non-planar)
• A unified view of different circuit classes (resistive, dynamic, AC)
• The realization that many "circuit" problems are really graph problems in disguise

Network theory does not add new physics to circuits. It reorganizes what we already know. It shows that the equations we solve are not arbitrary— they are determined by the structure of the connections. The components provide numbers; the graph provides structure; together they determine everything.