Explanation:

HTTP is stateless, meaning it does not maintain session information between requests. Each request is independent  .

HTTP follows a request–response model, where the client initiates a request and the server responds. That means communication is always client → server → client, not simultaneous in both directions.

Although newer versions (like HTTP/1.1 with pipelining or HTTP/2 multiplexing) allow more efficient handling, HTTP is still fundamentally half-duplex—it’s not full duplex because only one side can actively send at a time.

Why not the other options?

(A) Stateful – Incorrect. HTTP doesn’t preserve state between requests  .

(C) Bidirectional – Incorrect. While data does go back and forth, it's not spontaneous from both ends; the server only responds after a request.

(D) Full duplex– Incorrect. Only one side transmits at a time; true full duplex (simultaneous two-way communication) is seen in protocols like WebSocket, not HTTP.

So, HTTP is best described as a stateless, unidirectional (half-duplex) request–response protocol.

Answer: (B) unidirectional