Cell Relay Archive

[crystal <crystal@postme.net>]

Well, FWIK coaxial cable can never reach the potential of fiber optics.
I'm not familiar with STM-x, how many numbers do they go?

Bare fibre optics have a natural throughput of approx. 25,000 Mbps, and
this is assuming no multiplexing/Differentiate Wave Division
Multiplexing (DWDM)/Wave Division Multiplexing (WDM)/etc of the signals.

The only thing faster than fibre optics are super-conductors transceiver
ICs. You see the bottleneck is always the electronics, although the
electronic pulses are light speed - the microchip has to follow the
quartz clock (Mhz/Ghz) timing, because the bits must come into the
accumulators & registers in orderly words/byte/BCD etc. And since these
are electronic signals on silicon surfaces, things do get heated up.
It's just a smaller version of electricity flowing through a wire - the
free electrons "knock" against the fixed electrons in a closed circuit
from battery and back. The free electrons, driven by electrical
potential difference of the power supply/battery is the electical
signal.

Best Regards,
Crystal.

Fahad A Hoymany wrote:
>         Our network provider says they cannot provide us with an 
> optical interface to their network.  They can only give us an E3 
> electrical interface.   We want an STM-1 interface (155Mbps) but it 
> seems that Cisco does not have electrical interfaces for that 
> speedit only has optical interfaces for STM-1.  The questions are:
>  
> 1. Does it matter if you use electrical or optical STM-1 interface?

No difference in speed, but there is Electro-Magentic Interference (EMI)
in the form of magnetic fields around the electrical cable. This varies
according to the "shielding" of the cable. Aluminium-mylar or
aluminized-polyester (basically aluminium-plastic) foil reduces
high-frequency emissions. Braided copper reduces low-frequency
emissions.

EMI can cause temporary "spike" current/power surges in
electrical/electronic equipment in the immediate vincinity. Usu. this
means nothing, but if you have other sensitive electronic equipment
nearby (eg:medical MRI equipment/radio astronomy receiver/other
computers/etc), their performance would be disrupted. In a unshielded
computer, this would affect data processing (If severe. Like a lan cable
beside a bare mobo), esp. if incoming radio-comms modem link gets
disrupted (eg:aircraft comp to traffic controllers comp via
radio-signals modem/satellite-signals modem). Say, the my modem receives
frequency 123, but the EMI happens also to transmit frequency 123,
therefore my incoming data for any autopilot system is disrupted. That's
why onboard radar and/or GPS signals codes are secret. Although if the
GPS frequency is known, the GPS transceiver can still be disrupted -
just like bandit radio stations can broadcast on official frequencies
(abet limited geographic area). 

Fibre optic has no EMI. No EM radio receiver can tap its signals. Only a
direct cable splice and mini-ADC/DAC transceiver can do this.

> 2. If they give us an electrical interface to their network, does 
> that mean the actual signal over the fiber link (into their 
> network) is electrical?

Only the fibre optic transceiver IC (direct from the electronic
connectors) has the electronic signals. 

In a modem, the transceiver is a ADC/DAC (Analog-to-Digital Convertor,
and vice versa) between analog electical vis-a-vis digital electronic
signals. 

In fibre transceivers, it is digital LED/laser vis-a-vis digital
electronic signals.

> Is it even possible that an electrical signal be carried over a 
> fiber line?
 
Not possible. Howbeit, the light signals may be analog (eg:intensity or
changing from one wavelenght to another continously without pulsing) or
digital (pulses/sudden change in wavelength).