PPP
Point-to-Point ProtocolPPP is the handshake for a link with only two ends. Where Ethernet addresses many stations on a shared medium, PPP frames traffic over a single dedicated link, negotiates the options both sides will use, and authenticates the far end before any data flows.
PPP brings up a two-ended link and agrees its terms — framing, authentication and addressing in one negotiation.
How it works
Nothing is assumed: every option is negotiated, identity is checked, and only then does IP get an address. A Terminate request or a missed keepalive brings the whole link back down.
What PPP does, step by step
The Link Control Protocol negotiates frame size, authentication and keepalives until both ends agree.
CHAP challenges the far end with a nonce it must hash with the shared secret; PAP sends a password and is obsolete.
A network control protocol — IPCP for IPv4 — assigns addresses and options for each protocol to be carried.
Frames flow until either side sends a terminate request or the echo keepalives stop being answered.
In cybersecurity
PPP was the first link layer to authenticate before carrying traffic — but the mechanisms it shipped with are old, and some are broken outright.
PAP sends the password in the clear over the link; CHAP or EAP is the minimum.
It resists replay but not an offline crack of a weak secret — length still matters.
Its strength reduces to a single DES key; treat any tunnel relying on it as plaintext.
PPP authenticates the link but doesn’t encrypt it — wrap it in IPsec or use a modern VPN.
A rogue concentrator on the segment can answer discovery and become the far end.
ISP PPPoE credentials sit in router config in plain text more often than anyone would like.