Payment Systems: Magstripe, Tokenization, NFC and EMV

The fundamentals: how a transaction is verified and processed against the terminal, why the APDU protocol is not encrypted, and what protects the transaction instead.

The laboratory: how it is organized, how to navigate it, and which tool suits each payment technology.

Track encoding and the security thinking of the era, then MagSpoof, part by part, and the unsigned field on the stripe that tells a terminal to prefer the chip.

The APDU protocol end to end: packet generation, commands, responses, and the structure at the core of every transaction.

Contact transactions and the advanced APDU mechanisms behind them: the certificate chain a card carries, offline data authentication, and the cryptogram only the issuer can check.

The CVM List in tag 8E, how a terminal walks it to choose a method, what tag 9F34 records afterward, and the man-in-the-middle that let a terminal believe a PIN was verified while the card was never asked for one.

The part of EMV that decides rather than authenticates: the TVR in tag 95, Terminal and Issuer Action Codes, the zero floor limit, and Visa's published action codes decoded bit by bit, then the 2026 result showing one of those bits was never bound to anything.

NFC transactions in detail: APDU exchanges, cryptogram analysis, and cardholder verification methods.

The opposite of static magstripe data: a token is a different, independently valid card number, not an encrypted or seeded one, and how digital wallets implement it.

DPAN versus FPAN, the Token Service Provider role, provisioning, and where the keys live, then the 2021 research that paid from locked phones by flipping one byte, and the signed field that said so and was never read.

EMVCo QR, SPEI, CoDi, Pix and FedNow: static versus dynamic codes, settlement finality, and the fraud that followed the money out of the card rails.

Scoping, segmentation and what an assessor looks for, framed for someone attacking the environment rather than documenting it.

Reversing transactions to understand attacks seen in the wild, and building proofs of concept for each.

Tag 9F37 is the terminal's only contribution of freshness, and some terminals supplied a counter. What that made possible, why it is neither a replay nor a clone, and how to measure a terminal for yourself.

Why tapping stops after five times and asks for a PIN, where that counter really lives, and two pieces of published research that got past it: one resetting the counter without ever knowing a PIN, one making a plastic card present as a phone to an ordinary unattended terminal.

Relaying APDU data locally, over the internet and by MQTT, and what actually limits how much delay a relay can get away with. The countermeasure protocols themselves are module 17's job.

Mastercard's Relay Resistance Protocol, Visa's very different Level 1 approach, and where formal verification broke both. The sequel to module 16.

Commercial phones as contactless terminals: where PCI CPoC and MPoC sit, attestation, and what breaks when the terminal is an app on untrusted hardware.

ISO 8583, switches and acquirer hosts, HSMs, DUKPT key derivation, ISO 9564 PIN block formats and key ceremonies: the other half of the rail.

Anyone can buy a certified EMV reader now, and some of them take a payment with no network. A protocol-level walkthrough of four 2025 attacks: a loyalty protocol standing in for identity, a status word misread for years, and a signature over the copy nobody compares.

The card signs an amount it cannot verify, and the message that carries it restates that amount in a second place nobody signed. Following one cryptogram out through field 55 to the issuer and back as an ARPC, and a real message that admits nineteen different readings.

Three strands of original work that are not in the course yet, written as questions rather than results, each one saying plainly what has been demonstrated and what has only ever run in a unit test. The module is expected to change.