What Are Two Common Media Used In Networks Choose Two

You're staring at a multiple-choice question on a practice exam. "What are two common media used in networks? So choose two. " Your finger hovers over the options. Twisted pair. Fiber. Coaxial. Wireless. Worth adding: satellite. They all sound plausible. And that's the problem — they are all real. But only two are the answer the exam wants Small thing, real impact..

This changes depending on context. Keep that in mind.

Been there. Let's clear this up once and for all The details matter here..

What Are Network Media Anyway

Network media is just a fancy term for the physical stuff that carries signals between devices. Copper wire. Glass fiber. Radio waves. Even light pulses through the air, if you want to get exotic. The signal doesn't care what you call it — it just needs a path Still holds up..

In practice, when people say "network media" in a certification or enterprise context, they mean the wired physical layer. Practically speaking, the cables you terminate, label, and pull through conduit. Wireless is a whole separate domain with its own standards, spectrum rules, and headaches.

So right away, we can narrow the field. The question isn't asking about 802.Because of that, 11, Bluetooth, or microwave links. It's asking about the two cable types you'll actually terminate on a daily basis.

The Two Answers Every Exam Wants

Twisted pair copper (UTP/STP)

This is the default. The blue (or gray, or white) cable coming out of the wall jack behind your desk. Think about it: category 5e, 6, 6A, 8 — they're all twisted pair. In practice, four pairs of copper wires, twisted together at specific rates to cancel electromagnetic interference. Think about it: the twisting is the whole trick. Without it, crosstalk would murder your signal at anything above a few megabits.

Not obvious, but once you see it — you'll see it everywhere.

Unshielded twisted pair (UTP) owns the LAN. Shielded twisted pair (STP) exists for noisy environments — factory floors, near heavy machinery, alongside power lines — but it's stiffer, pricier, and requires grounded connectors and patch panels. It's cheap, flexible, terminates with an RJ45 connector in thirty seconds once you've done it a few hundred times, and handles 1 Gbps to 10 Gbps over 100 meters without breaking a sweat. Most offices never need it.

Fiber optic cable

The other answer. Glass or plastic cores carrying light instead of electricity. No EMI issues. Here's the thing — no crosstalk. No grounding nightmares. Distance limits measured in kilometers, not meters. Single-mode for long haul (10 km, 40 km, 100 km+). Multimode for shorter runs inside buildings (300–550 meters at 10 Gbps, depending on grade) Most people skip this — try not to..

Termination used to be a dark art — epoxy, polishing, microscopes. But for backbone links, data center interconnects, and anything between buildings, it's not optional. Now you've got mechanical splice connectors and fusion splicers that make it approachable. Still slower than crimping an RJ45. Still more expensive per drop. It's the only thing that works.

Why These Two Dominate

Everything else is niche. Still, coaxial? In practice, you'll see it on a cable modem handoff or an old CCTV run. Maybe a legacy Thinnet segment in a museum. In practice, it's not a general-purpose LAN media anymore. Wireless? Think about it: essential, but it's not cable media — different physics, different standards, different troubleshooting. Satellite? Now, latency makes it a last resort. Powerline? Unreliable. Free-space optics? Weather-dependent No workaround needed..

Twisted pair and fiber cover 99% of wired infrastructure. One for the horizontal run to the endpoint. One for the vertical backbone and long-distance links. That said, that's the architecture. Has been for twenty years. Will be for twenty more Worth keeping that in mind..

How They Actually Work in Practice

Twisted pair — the daily driver

You pull Cat6 from a patch panel to a wall jack. Because of that, test with a cable certifier — not just a continuity tester, a certifier that verifies NEXT, PSNEXT, ACR-F, return loss, propagation delay. 5 Gbps and 5 Gbps work on Cat5e and Cat6 respectively. 2.If it passes, you're good for 1 Gbps all day. Punch down both ends with a 110 block tool. 10 Gbps needs Cat6A (or Cat6 with strict length limits and alien crosstalk mitigation) And that's really what it comes down to. And it works..

Real talk: most "Cat6" cable sold cheap online is actually Cat5e with a Cat6 jacket. Terminate to T568B unless your org mandates A. Don't zip-tie bundles tight. Here's the thing — keep bend radius loose — four times the cable diameter minimum. Buy from a real distributor. Now, don't mix them in the same run. Don't run parallel to power cables closer than 12 inches (cross at 90 degrees if you must).

Fiber — the heavy lifter

You're connecting two IDFs across a campus. You fusion splice pigtails onto each fiber, slide them into a splice tray in a rack-mount enclosure, patch to your switch SFP+ modules with LC duplex jumpers. Run an OTDR trace to verify splice loss under 0.5 dB. That said, single-mode OS2, 12-strand, armored outdoor rated. But 3 dB per splice, connector loss under 0. Document every strand end-to-end.

Multimode inside the building? Which means one-click cleaners are cheap. Plus, oM4 or OM5. 50/125 micron core. VCSEL sources at 850 nm. MPO/MTP connectors for high-density trunk cables breaking out to LC duplex at the switch. Which means clean every connector every time — one speck of dust on a 9-micron core kills the link. Use them Easy to understand, harder to ignore..

Common Mistakes People Make

Treating all twisted pair the same. Cat5e, Cat6, Cat6A — they look identical. They are not. The twist rates differ. The pair separators differ. The jacket materials differ. The alien crosstalk performance differs. If you terminate Cat6A like Cat6 (untwisting too much at the jack), you just bought expensive Cat5e Took long enough..

Mixing single-mode and multimode. Seen it. SFP module says 1310 nm single-mode. Patch cable is orange multimode. Link light never comes up. Worse — you can sometimes get a link with mismatched fiber and modules at short distance, but it's unstable and you'll waste hours troubleshooting Most people skip this — try not to..

Ignoring bend radius on fiber. Copper kinks. Fiber shatters microscopically. Microbends add loss. Macrobends leak light. That 90-degree turn behind the rack? Just added 2 dB loss. Your 10 km link budget just became 6 km.

Skipping certification testing. A $50 continuity tester tells you the pins connect. A $8,000 certifier tells you the performance meets spec. Guess which one saves you from "intermittent slowness" tickets six months later.

Labeling nothing. Or labeling with a Sharpie that rubs off in a week

Best‑Practice Checklist for a Trouble‑Free Deployment

1. Document every step in a living diagram – Use a network‑drawing tool that can export to PDF and embed version numbers. Include cable part numbers, jacket ratings, and the exact patch‑panel ports that terminate each run. When a future technician opens the rack, the diagram should read like a map, not a puzzle.

2. Adopt a standardized labeling scheme – Instead of a handwritten tag that fades, employ heat‑shrink sleeves with laser‑etched text or QR codes that link to a central asset‑management database. Group cables by function (e.g., “Core‑Uplink‑01‑LC”) rather than by individual workstation, which simplifies troubleshooting and reduces the chance of mis‑patching Less friction, more output..

3. Maintain a spare‑parts inventory – Keep a small stock of identical patch cords, keystone jacks, and spare fiber pigtails on hand. When a link drops unexpectedly, swapping a known‑good component is often faster than tracing the fault through a maze of cables Worth knowing..

4. Schedule periodic performance verification – Even after a flawless certification, environmental changes (temperature swings, added power sources, new equipment) can erode margins. A quarterly run of a portable certifier on a random sample of links catches drift before it becomes a service‑impacting issue.

5. Plan for future bandwidth growth – When laying new runs, oversize the conduit diameter and leave extra fiber strands in the trunk. A 24‑strand OM5 cable can later be broken out into 40‑GbE or 100‑GbE connections without pulling new fiber through the same pathway.

6. Secure cable pathways against accidental damage – Use sturdy cable trays with reinforced corners, and install protective conduit where cables pass through walls or ceilings. In high‑traffic zones, consider conduit sleeves with a larger bend radius to accommodate future upgrades without stressing the existing wiring Worth keeping that in mind..

7. Integrate cable management into the change‑control process – Any new patch, relocation, or addition should trigger a review of the existing pathway map. A simple checklist — “Is there free space? Are bend‑radius limits respected? Are labels still legible?” — prevents ad‑hoc modifications that degrade performance over time.

8. Educate the entire team – Conduct brief “cable‑care” workshops that cover proper handling of fiber, the importance of clean connectors, and the risks of mixing cable categories. When every stakeholder understands the rationale behind the standards, compliance becomes a shared responsibility rather than a compliance checkbox.

Conclusion

A well‑engineered cabling infrastructure is the silent backbone of any high‑performance network. Here's the thing — future‑proofing isn’t an afterthought — it’s a design decision that saves time, money, and frustration when the next generation of applications demands more bandwidth. Treat your cabling as a living system: monitor it, maintain it, and upgrade it deliberately. By respecting the electrical and mechanical limits of each cable type, rigorously documenting every connection, and investing in quality testing and labeling, organizations eliminate the hidden sources of latency, downtime, and costly re‑work. When done correctly, the network will deliver the speed, reliability, and scalability that modern workloads require — without the need for constant troubleshooting or emergency patches That's the part that actually makes a difference..