AMI
Understanding Alternate Mark Inversion (AMI)
In the 1960s, the telephone company tried to send millions of digital 1s and 0s across the country using massive copper cables. They quickly ran into a terrifying physics problem: sending too many '1s' in a row caused a massive backup of electricity that melted the equipment. To fix this, they invented a genius mathematical code called Alternate Mark Inversion (AMI).
The DC Voltage Nightmare
In a standard computer, a '0' is zero volts, and a '1' is 5 volts.
If you send a file containing ten thousand '1s' in a row down a long copper wire, you are essentially just hooking the wire straight to a massive battery. This creates a massive, continuous wave of raw electricity (a DC component). This raw electricity cannot pass through the magnetic transformers used in the telecom network. It hits the transformer, turns violently into heat, and destroys the telephone pole equipment.
The Alternating Miracle
AMI completely solves this problem by using three different voltage levels.
- A digital '0' is sent as 0 Volts (the wire is silent).
- The first digital '1' is sent as a positive pulse (+3 Volts).
- The very next digital '1' is sent as a negative pulse (-3 Volts).
By forcing every '1' to flip upside down, the electricity perfectly balances itself out. The positive pulses and negative pulses completely cancel each other. The average DC voltage of the copper wire remains exactly zero, allowing the massive telecom network to run flawlessly for decades without melting.
Key Equations
Alternate Mark Inversion (AMI) is a foundational, pseudoternary physical-layer line coding protocol historically utilized in high-speed digital telecommunications (specifically the North American T1 and European...
Key specifications:
1 a | 0 V | 3 V | -3 V | 5 v
Power: P(dBm) = 10log(PmW), 0dBm = 1mW
Comparison
| Aspect | AMI Spec | Typical Range | Impact | Design Note |
|---|---|---|---|---|
| Primary function | In primitive binary transmission (Non-Re... | Application-dep. | Critical | Verify in sim |
| Operating range | This massive DC component creates catast... | Application-dep. | Critical | Verify in sim |
| Performance | AMI mathematically solves this physics f... | Application-dep. | Critical | Verify in sim |
| Integration | A digital '0' is transmitted as exactly... | Application-dep. | Critical | Verify in sim |
| Trade-off | A digital '1' is transmitted as a high v... | Application-dep. | Critical | Verify in sim |
Frequently Asked Questions
What happens if there are too many zeros?
That is the single biggest flaw of AMI. If you send a file with thousands of '0s' in a row, the wire goes completely silent (0 Volts). The computers receiving the signal lose the physical timing clock and 'lose their place' in the data stream, causing the network to crash. To prevent this, the telecom company invented 'B8ZS' (Bipolar with 8-Zero Substitution)—a secret code that automatically injects fake, intentional errors into the wire anytime it sees eight zeros in a row, forcing the clock to keep ticking.
Is AMI still used in modern fiber optics?
No, it is strictly obsolete in fiber optics. AMI was specifically invented to solve the DC voltage problem in physical *copper* wires. Because fiber optic cables use pulses of light (photons), there is absolutely no electrical voltage, and therefore no DC baseline wander problem. Modern fiber optics use entirely different, vastly faster coding schemes (like 64b/66b or PAM4).
How does AMI catch errors?
By definition, the code is self-policing. Because the rule states that every '1' MUST alternate polarity (+, -, +, -), if the computer suddenly sees two positive pulses in a row (+, +), it knows with absolute mathematical certainty that the copper wire was struck by lightning or interference. This is called a 'Bipolar Violation', and it instantly triggers the telecom computer to flag the data packet as corrupt.