What a coincidence. I'm doing exactly that right now.
There are lots of ways to do this, but controlling the impedance is critical. The problem with using an unshielded ribbon cable is that it is hard to manage the impedance. It can be done, but finding a good cable supplier is hard, especially one that can guarantee the impedance. On the surface this looks like a good solution, but the difficulty of suppliers and stuff makes this less attractive.
Another way to do it is to use a shielded ribbon cable. I don't like this approach because the cable is expensive and terminating the shield is next to impossible.
The Flex-PCB is a good solution, but the NRE is high. If this solution is within your budget it might be the best thing.
A "quick and dirty and cheap" solution would be to use a couple of SATA cables! The impedance is right, they are cheap and widely available, and more importantly they work. Of course you will be limited in the number of signals so don't expect to get more than a PCIe x1 connection plus a reset signal (no power, JTAG, etc.).
For transmission of signals over very long cable runs in EMI-rich environments, the go-to standard in industrial applications is the current loop. As pointed out by @Haneef, Wikipedia also has a page on Digital Current Loops, albeit with hardly any useful details.
The source signal is converted to a regulated current drive, of a specified range, the most common being 4-20mA, and 10-50mA.
As the current in the loop will be constant throughout the loop, this mechanism has high EMI tolerance. The minimum current in the specification (4 or 10 mA in the above cases) also provides a way to detect cable breaks or remote transmitter failure.
There are standard 4-20mA and 10-50mA current loop driver ICs available from many of the major IC manufacturers, that can take analog or digital voltage inputs. These ICs are specified for a variety of signal bandwidths, so one can choose the least expensive one for the cable length, power supply voltage and desired bandwidth, for one's application.
That being said, a 1.5 meter cable run is not really considered a problem even for standard 5V-TTL voltage signals, easily handled by shielded cable, unless the cable passes right next to some truly massive EMI source, such as HVAC equipment, large switching systems, or transmission power line equipment.
For 3 digital signals, the kind of shielded twisted pair cable used for Ethernet, or for shielded USB cables, is commonly available and inexpensive. Each of the digital signals should be applied to one of a twisted pair of cables, with the other wire of the pair in each case being connected to signal ground. This attenuates common mode interference massively, over and above the protection provided by the shield ie conductive braid wrapped around such cables.
See this page for illustrations of types of shielded (and unshielded) twisted pair cables, thus:
Best Answer
STP I assume means screened/shielded twisted pair and if you want to measure current flowing down one of the conductors then do so at any point. "Condition" the signal where the measurement is made or do the conditioning half locally and half remotely. The minimum amount of conditioning to be done locally is to amplify the shunt signal so that the wires from the amplified shunt conditioning circuit do not pick-up excessive extraneous EMI on route to the final circuit that monitors the shunt current.