One of the less glamorous challenges of working from home is making sure the internet connection can carry a real workday. It is not just the video and voice calls. It is screen sharing, transferring large files, cloud applications, and always-on services such as OneDrive quietly synchronizing in the background. Add the extra distance and processing that can come with a corporate VPN, and a connection that seems acceptable for casual browsing can begin to feel painfully inadequate.

Like many remote workers, I had been doing my best to improve the connection I already had. For me, 5G home internet remains one of the most economical and relocation-friendly options available, and I use T-Mobile. The gateway is easy to move and set up, although service eligibility still needs to be confirmed for a new address.

Then I relocated to an apartment in San Antonio, and the experience changed.

When an adequate connection stopped being adequate

At the new apartment, my download speed was slow enough to interfere with work. Calls were less dependable. Cloud work felt constrained. Even ordinary tasks became exercises in waiting and wondering whether the connection would hold.

I started with the usual troubleshooting. I connected my laptop directly to the gateway with an Ethernet cable so Wi-Fi inside the apartment would not muddy the test. I moved the gateway from one window to another, trying windows that faced different directions. I let the connection settle, ran more tests, and hoped I would find the one spot where everything clicked.

Nothing made enough of a difference.

I seriously considered going back to cable or fiber. Either one likely would have solved the immediate performance problem. The complication was cost. My home internet and mobile lines were bundled, and canceling the 5G service could have meant restructuring the plan at a higher price, then adding the cost of a new internet service on top of it.

Before taking on another recurring bill, I decided to try one more thing.

My last experiment: a compatible external antenna

I purchased a 4x4 MIMO omnidirectional external antenna designed for 4G LTE and 5G frequencies. The price at the time was $49.95. Most of my initial concern was practical: if the antenna worked best outside, how would I route four cables through a window or patio door without creating a new problem?

For the first round of testing, I skipped that question. I kept the antenna indoors, connected its four leads to the matching ports on my compatible gateway, positioned it near a window, and ran a controlled comparison.

The improvement was not subtle.

‍

‍

The before-and-after results

Before adding the antenna, six tests produced an average download speed of 8.36 Mbps. After adding it, eight tests averaged 184.78 Mbps. That is about 22.1 times faster.

The download range also became far more useful. Before the antenna, results ranged from 2.9 to 11.7 Mbps. Afterward, they ranged from 158 to 210 Mbps.

Average ping dropped from 85.8 milliseconds to 42.3 milliseconds, an improvement of about 51 percent. For interactive work, that matters. Raw download speed gets most of the attention, but latency has a direct effect on how responsive calls, screen sharing, remote sessions, and cloud applications feel.

One measurement did move in the wrong direction: average upload speed fell from 19.00 to 10.87 Mbps. That was still usable for my video meetings, but it is an important reminder that an antenna does not automatically improve every metric. If your work involves frequent large uploads, live broadcasting, or off-site backups, upload performance deserves its own careful testing.

Test conditions: The before average is based on six tests and the after average on eight. All tests used the same Brightspeed Speedtest server in San Marcos, Texas. The antenna remained indoors for this comparison.

‍

Why an antenna can help - and why results will vary

Moving the gateway around the apartment was still the right first step. T-Mobile itself recommends placing a compatible external antenna in a window when the gateway cannot be located in an ideal spot. In my apartment, however, repositioning the gateway alone did not create a stable enough connection. The external antenna gave me more flexibility in where the cellular signal could be captured while keeping the gateway and wired equipment where they were practical to use.

That does not mean my numbers are a forecast for someone else's home. Cellular performance can change with distance and direction to the tower, building materials, the frequency bands in use, network congestion, local mobile traffic, equipment, and weather. Apartment units in the same building can behave differently simply because their windows face different directions.

The useful lesson is not "buy this and expect 22 times the speed." It is this: before replacing a reasonably priced 5G service with a more expensive monthly option, test whether the existing cellular link can be improved.

A practical way to test your own setup

Build a baseline. Run several tests at different times of day. Record download speed, upload speed, and ping. If your gateway exposes cellular signal metrics, record those too.

Remove your indoor Wi-Fi as a variable. Connect the test computer to the gateway with Ethernet when possible.

Try placement first. Test different windows, heights, and orientations. Give the gateway time to reconnect and settle after each move.

Confirm compatibility before buying. Gateway models differ. Check that your device supports an external antenna, that the connectors and frequency range match, and that you understand the manufacturer's setup instructions.

Change one thing at a time. Keep the test device, server, and general test conditions consistent so you know what caused the difference.

Retest when the network is busy. A good result on a quiet weekday morning may not reflect performance during evening congestion.

Judge the work experience, not only the speed test. Track call stability, screen-sharing quality, file-transfer time, VPN performance, and cloud-sync behavior for several days.

Home connectivity is part of the workplace

This small experiment also carries a broader lesson for organizations with remote or hybrid employees. Home internet is now part of the workplace technology stack, even though employers do not directly control it. A person's connection may be supporting a conference call, a VPN tunnel, a large presentation download, and background synchronization at the same time.

That makes connectivity support a productivity and resilience issue, not merely a personal convenience. Employers can help by setting realistic bandwidth expectations, offering flexible equipment or connectivity stipends, and having IT teams examine whether VPN routing is adding avoidable latency. Microsoft, for example, recommends carefully configured split tunneling for key Microsoft 365 traffic in appropriate enterprise environments; that is an IT and security decision, but it illustrates how network architecture can shape the remote employee experience.

A $50 test before another monthly bill

In my case, a one-time antenna purchase moved the connection from marginal and disruptive to well beyond what I expected. It did not make cellular internet perfectly predictable, and it did not improve upload speed. It did, however, give me the download headroom and lower latency I needed for a much better work-from-home experience.

Your results may be modest, dramatic, or nonexistent. That is the nature of cellular service. But if you have already tried a wired connection and several gateway locations, and you are close to abandoning an otherwise economical 5G plan, a compatible external antenna may be the next experiment worth running.

Sources and product notes

The measurements in this post come from my own before-and-after testing summarized in the accompanying presentation. The antenna tested was listed as a 4x4 MIMO omnidirectional wide-band 600-6000 MHz 4G LTE/5G external antenna (ASIN B0H68MLSJ7), priced at $49.95 when purchased. Pricing and availability can change.

T-Mobile: G4AR and G4SE gateway setup and external antenna guidance

T-Mobile: Transferring Home Internet service to a new address

Microsoft Support: Changing OneDrive upload and download rates

Microsoft Learn: VPN split tunneling for Microsoft 365

After moving to San Antonio, my T-Mobile 5G Home Internet averaged just 8.36 Mbps down. A roughly $50 indoor external antenna raised that average to 184.78 Mbps and cut ping by about half. The upload result was less impressive, and cellular performance always varies, but the test reinforced a useful lesson: before adding another monthly bill, measure whether the connection you already have can be improved.

‍