[Network] Search for UniFi Alternative | Is there one?

I have had full UniFi network system at my home for almost 3 and half years. It has been working well, but as a sporadic, casual network hobbyist, I wonder if there is a better alternative. This is by no means a comprehensive list, but if you are on the market and wondering what other options you may have besides UniFi, I hope this helps you provide at least starting point for your search.

Update: 12/26/2024 – Edit Aruba Instant section as in the end I’ve ended up purchasing them as well.

Background

My whole network hobby started from a necessity.

A little over 5 years go, we’ve moved to old 1950’s built, 2 story, approximately 3800 sqft home. Back then I had to work from home at least couple times a week ,which were quite dependent on the constant data downloading. 

There was no ethernet jack anywhere in the house. Fortunately, the cable were present throughout home, so we put cable modem at the center of house with a Netgear wireless router attached to it. Not surprisingly, WiFi signal did not reach to my walk-in basement office room.  So I discovered mesh network system, which was something I never needed while living in apartment.

Era of Mesh Systems

Orbi

As a WiFi novice, I simply searched for the highest rated mesh WiFi system across multiple sites over the internet. That was Orbi almost unequivocally. So I purchased the flagship version of the Orbi.

Although I was able to get a setup to get decent WiFi internet speed test in all the crucial area of home, its stability made essentially unusable.

Old review

Eero Pro

Eventually, I gave up on Orbi and decided to move on to something else. This time, I cared less about the highest speed, which again was Orbi back then. But instead I search for the most stable system. Even though the specification looked underwhelming, it actually served our family very well. During the seven months I had this set up in our house, I got no complaints from my family regarding to WiFi.

Eero Pro Review from 2020

Asus AiMesh

Not surprisingly, I like techs. This was around the time earliest models of WiFi 6 routers start to show up on the market. Also, at this point I got cat 6 drops in multiple rooms in my home as well as we got 1 Gbps down speed. So I wanted to be able to utilize that, and Eero did not have the enough spec for it.

So I went for what looked to be very clever, perhaps innovative design, AiMesh system by ASUS. This allows one Asus router to be the router, and other routers to become either wired access point or mesh node. All integrated and controlled in a single control panel.

I was able to get WiFi 6 client and routers. See the good speed test and iPerf test that were not possible with Eero. However, even with wired backhaul AiMesh degraded performance of each node/router throughput compared to stand alone. Also, despite it was usable, they still suffered relatively stability issue where I had to reboot the system every a few days. The software was buggy and many features that sounded cool were something I just had to turn on to make it work reasonably i.e. really those features weren’t there for me.

AiMesh with Ax11000 review

Era of UniFi

Since then, it’s been over 3 years. The most recent UniFi set up consists of 3 U6 Pro access points, 1 SHD access point, and U6 mesh. All wired. On my server rack, I have Unifi Dream Machine Pro, USW-Pro 24 POE, USW-Aggregation, USW-Flex-XG , along with 3 USW Flex Mini. 

However, despite being the most recent addition to the network, recently the U6 Mesh died with merely a little over a year and half of use. The unit felt quite hot from the day 1, so I was worried about this device may not last long, but it was rather faster life span than I thought. 

Since it’s already out of warranty, my options were order replacement, buy upgrade such as U6 Enterprise. Fortunately, I still have corner to corner WiFi coverage throughout my home, just not at max speed everywhere, but still pretty much every where over 200 Mbps. So I could also wait for WiFi 7 version of access point to release. 

But then the third option came to my mind. What if I start to look for entirely new Wireless network set up?

Despite UniFi had served me very well so far, it was not a problem free 3 and half years. There are times my wired and wireless network gave me issues that felt to be arising from UniFi network. Each time I upgrade to new firmware, I worry for possibility of something could break on my network though this seems much less frequent lately.

Some practical impacts includes roaming issue that constantly drops WiFi call in the same area of house. This happens at the spot where one of the Access point is located within 10 ft.  Sometimes, new IoT device setup requiring way more troubleshooting than I had imagined.

Then there are features that looks promising, or interesting as network hobbyist but they never work, or can’t find actual documentation what they do. 

 Search Criteria

So I thought this maybe a good time for me to at least do some research about alternative options, especially since I now have a little bit more of basic network understanding when compared to 3 and half years ago. I can at least make an educated guess for what it would be the new learning curve for me if I were to go with new system.

Despite something to desire, I have been fairly happy with my current UniFi WiFi setup.This makes my search criteria for alternative are much more strict. 

So I have listed my personal new ideal WiFi system’s requirement in order of priorities:

1. Stability

2. Scalability/Modularity

3. No mandatory subscription

4. Local controller system 

5. Speed

Stability 

Stability refers to uninterrupted WiFi connection. In my opinion, the ideal, WiFi network should be like air we breathe. Once set up, we forget about it other than the times I want to play around with new network settings as a hobby.

In larger scale, the stable system should require essentially no to very seldom rebooting. The system after weeks or months of running should perform as well as it did on the day of reboot.

At the level of daily use, we should have uninterrupted streaming, gaming, zoom meeting and WiFi call, all at once. We are just single family with 4 people. 

Previously, I misinterpreted stall on streaming, video chat or Voice over Internet as speed or throughput issue. But Netflix requires just over 15 Mbps for its 4K streaming. Zoom video calling recommends only 4 Mbps for 1080p HD even on group meeting. 

With 2 Gbps down/200 Mbps up asymmetric service and iPerf test of 100 Mbps every corner of my home, most part over 400 Mbps, these issues cannot not related to throughput. They are usually related to latency, network congestion. So I categorize these issues as “stability” issue.

 I also categorize poor roaming in multiple access point system as stability issue.

 Technically, these classification maybe arbituarity but I know they are usually separate from conventional internet speed test or local network iPerf test result. 

Since these are the issue, I am experiencing with my current UniFi system, the new system if any must solve it.  If these issues improve, then I call stability is improve, and that’s actually day to day practical end-user experience improvement.  It’s far more important than high iPerf number or Internet speed test number.

Scalability

Modularity or scalability may not be as important to many home users, but if you are watching this video, I think you are probably like myself who are network enthusiast/hobbiest who maybe thinking or at least like to think about optimizing or upgrading your home network. 

In such situation, I think modularity and scalability of the network system is important. Being able to upgrade, swap or add one piece at a time allows me to try out a cutting edge technology, troubleshoot, and improve overall network experience without overhauling the entire network each time.  

In technology, I believe there is no true future proof. It’s all relative because everything gets better in fairly short time. So if someone is looking for the latest and the greatest, I think one way to look at future proof system is it’s a modular/scalable system where you can add/swap each piece and continue to evolve the system as you desire.

Local controller system

This is sort of continuation of no subscription.

Technically, I am ok with cloud based controller so long as there is local controller option. Some wants to make sure network controller is accessible even during internet is down, but I personally don’t see myself needing to change WiFi setting in such urgency. So long as the WiFi network continues to work during internet down with existing setting, I have no issue with it.

However, the concerned  I have when a system does not offer local controller option is the potential risk of when the manufacture decides to stop providing the cloud service. If that happens, then we’d lose control to our device permanently at that point. 

Realistically speaking, the chance of this happening when purchasing from major brand is very low, but end of life devices and losing their support can be a real thing. Although as some who likes newer technologies, I’d most likely have upgraded to latest device well before that anyways, but I still don’t like to think for losing something after purchasing. 

Speed

Finally, the last but still important criteria category for me is the “speed”.

This is actually an interesting topic of its own. Before I knew much about the network, I had thought network performance is best depicted by internet speed test and iPerf type speed test. The bigger the number I can get, the better the network is. 

However, this did not necessary translated to end user experience, and I’ve learned that fairly early stage when started playing with WiFi system. My original consumer grade mesh network system setup with Orbi had 4 times speed test numbers compared to Eero mesh in the same area of my house, but Orbi had stability issue while Eero didn’t. 

My family members who knows nothing about WiFi had zero complain while I had Eero set up but I got constant complaints with Orbi and AiMesh. Yet, speed test numbers were all higher than Eero.

Another main reason why I am putting this on the lowest criteria is this is a factor that has the least likely variation because they usually follow each device specification well. Don’t get me wrong, there are difference amongst devices with relatively similar specs. Many company brags about their proprietary, innovative technology to distinguish themselves. There are many online sites and Youtube videos to show the difference amongst devices with similar specs. And I myself have tested some in the past and definitely noticed that there are real difference, yet I feel such difference may be relatively small, and may not necessary translates to end-user experience as much.

So specifically, what are the access point specifications that I am looking at?

Wi-Fi standard.

WiFi 6E is here. I only have two client supports it at the moment, but one is wired at 10Gbps. If the price is right, I would not mind it but if if it costs a premium, I’d rather pay for that to WiFi 7. It’s important to note that WiFi 7 standard is technically not yet finalized, so there is still some chance things can change even if there were WiFi 7 access points available right now.

MIMO

Standard MIMO option to consider are 2×2 vs. 4×4. Practically speaking, most clients I own are 2×2 except old MacBook Pro has 3×3 WiFi 5, which could be irrelevant.

Technically, if we have 4×4 MU-MIMO capable AP and MU-MIMO capable clients, I should be able to get two 2×2 MIMO clients connecting simultaneously at their max speed. 

However, I don’t know if Apple products supports it. I read some websites says yes, but I can’t see official Apple statement. 

In fact, I’ve tried iPerf test across two Wirelessly connected Apple devices and individually down stream iPerf simutaneoutaneously across MU-MIMO capable UniFi access point, and the speed cut in half. So I think iOS devices do not support  MU-MIMO. 

Therefore, I am ok with 2×2 option. 

Channel Width

Channel Width affects throughput proportionally. The main question here is do I need 160 MHz?

In 5GHz band, 160 MHz is a possible option. If that works, then that alone can get us 2x throughput than typical 80 MHz channel width setting at home network. 

There are couple challenges here. The first is I do not believe any Apple product, which is our primary house hold WiFi clients support it. Second, this channel setting will require to use Dynamic Frequency Selection channels, which in my area can actually trigger and be cause of network instability. Thus, I won’t be able to use 160MHz in 5GHz band at my home, and therefore, not important.

If you are Android users, single AP setting, and no concern of DFS strike, 160 MHz support maybe an important criteria.

So basically, from the speed aspect, my requirement is relatively low.

I need to make sure the platform has access point with at least WiFi 6, 2×2 MIMO and 80 MHz channel width capable on 5 GHz. 

Though this is my practical need, as a tech enthusiast, it is nice to know the platform supports newer technologies in timely fashion. So I won’t be needing it right at this moment, it would be a plus to see WiFi 6E support, and perhaps promise of WiFi 7 support as soon as the standard is finalized.

UniFi Alternatives

Now that I have defined my search criteria, I’ve started to look at my options based on the category of devices from consumer grade all way up to enterprise grades.

Consumer Grade Gears

Starting with consumer graded gear.

One major benefit of consumer gears are they adopt newer technologies much faster. There are already WiFi 7 routers and mesh systems available in this category of WiFi systems.

This includes major consumer brands, Netgear Orbi,  Linksys Velop, Amazon Eero, Asus AiMesh compatible routers and so on.

One thing I found interesting is that Eero now looks to allow newer Eero mixed with older Eero devices. Previously one of my main issue with consumer mesh system were even from the same brand, you had to buy entire new mesh system. This can add up cost very quickly as you were essentially replacing everything. Eero states when mixed with older device, you can only get the older device specification level of performance from the over all system. So this feels half scalable, and certainly improvement from the past, but still not quite the same as fully scalable system.

For this reason, if I were to choose a product in this category, I’d go with AiMesh system. AiMesh allows real scalability similar to prosumer or enterprise products. AiMesh allows us to use older router an access point or mesh node of an integrated system while still getting full power of the new router.

For example, if I still had AiMesh setup with WiFi 6, I could have just bought one WiFi 7 router and use that as a router, and old WiFi 6 router becomes access point or wireless mesh node.

If WiFi 7 device is connecting to WiFi 7 router, it gets WiFi 7 benefit. It is only when the client associated with WiFi 6 AP/node then it gets down graded to WiFi 6 connection. So AiMesh allows someone like myself to continue adding/replacing a piece at a time and keep up (play) with the latest technology or maybe just adding cheaper node purpose routers to expand WiFi coverage without down grading the whole WiFi network as in Eero’s case.

Source: https://www.asus.com/microsite/aimesh/en/index.html

However, the modularity and scalability are not at the level of prosumer/enterprise gears. When knowing that I only need access point functionality, paying for full router price seems to be not the best way to spend money. I’d rather spend the same money to get better access point dedicated unit, with more functionality as AP. Another example of lacking scalability is even with AiMesh, it still ties me to only Asus AiMesh compatible products. In contrast, as I am doing right now, with Access Points completely separated from the remaining network, despite I have everything UniFi at the moment, there is nothing prevents from me to switch Access Points to entirely new brands, or perhaps just have couple APs with new brands and keep remaining UniFi etc. There are just much more options.

But, the main reason I have decided to not go back to this category is that I am still worried about consumer gear’s stability. I do not want to reboot system every week or month to get the system’s performance back.

Prosumer Grade Gears

Although some says UniFi gears are good for business setting, I often hear they fall under prosumer gears category, which I agree.

In this category, I found 3 primary contenders.

• Aruba Instant On
• Meraki Go
• TP Link Omada.

The interesting part of these 3 platforms are where they are coming from.

I believe TP-link is known for consumer graded network gears. Omada series is their step into small business category. 

While Aruba and Meraki are well known enterprise network brand. Instant on and Meraki Go are their prosumer/small business grade versions and unlike their enterprise counter parts, both requires no license or subscription. 

Having said that, it is important to note that Aruba and Meraki enterprise gears are completely separate from Aruba Instant On or Meraki Go products. So you can’t use Instant On gear on enterprise network or vice versa. I have not researched deep enough to say what have been stripped off on prosumer models compared to enterprise models in these platforms from both hardware and software, but I would assume there are some significant changes.

Instant On & Meraki Go

I’ve also read on Reddit, it maybe possible Aruba Instant On uses the same hardware as the enterprise version but you just can never convert one from the another.

Whether the same internal hardware is true or not aside, both Meraki Go and Aruba Instant On certainly lacks catalogue selection when compared to their enterprise listing, but also when compared to UniFi. 

Meraki Go’s top of the line is 2×2 WiFi 6, 1 Gbps Ethernet port Access Point. Aruba Instant On has 4×4 WiFi 6 Access Point with 2.5 Gbps ethernet port at its top of the line model. However, they still does not offer WiFi 6E. 

While this is not a deal breaker for me at the moment, it tells me the latest technology is not these product line’s top priority; hence, there are good chance, their WiFi 7 product shipping may also be significantly later than UniFi.

Also, in regarding to price for relatively similar models,  UniFi looks more attractive to me. 

Although I’ve read several users commenting they’ve moved from UniFi to Aruba Instant On, on the paper, I did not feel the enough potential for the switch.

Omada

Their top of the line product has WiFi 6E support with 10 Gbps uplink Ethernet port. When we look a the U6 Professional equivalent access point, it has 2.5 Gbps uplink as opposed to the U6 Pro’s 1 Gbps while Omada AP is cheaper. 

Furthermore, Omada offers what the real enterprise platforms are offering. That is besides the license and subscription free software based controller, users can purchase hardware controller and even subscribe for cloud based controller for added features. 

Though in this regard, one must be a bit skeptical for actual content. I have not researched enough, but just looking at their official website summary table, they list what I consider the most fundamental form of radio resource management such as auto channel selection, cell size adjustment. The feature I had hoped UniFi’s once called WiFi AI would have done for us. However, this is listed under cloud based controller feature. When we talk about the classic enterprise manufactures, we don’t need to pay for these, it’s free build in. 

Another thing that makes me skeptical is the fact, they list such basic feature as part of the list rather than more sophisiticated proprietary technology even if they were just market talk. 

On paper, Omada looks very appealing. 

This is my personal bias, but I had always viewed TP-Link as a consumer gear brand, and thought they were budget friendly product but never looked them as premium, quality first product.

One of UK blogger nicely goes in detail of comparing UniFi to Omada as a whole platform. If you are interested in Omada, I highly recommend you to check the blog out.

They states Omada as having slightly reputation for second-class product” which is concordant with my notion on the brand from consumer gear side. However, the author says part of it might be due to “lower price” point of TP-link maybe at least partly playing that reputation.

Their conclusion was generally suggest installing Unifi because aesthetic and user experience. The performance are compatible, but Omada has fewer bugs than UniFi. For those who are security conscious, be aware for TP-Link being Chinese company.

My personal impression of Omada just on paper is they can be compatible to UniFi, but I did not feel it is something unanimously viewed as superior product. 

Enterprise Gears

So unlike 3 and half years ago, this time I’ve decided to take a look at a level above, the actual enterprise market. At this level many require subscription and license, so that alone narrowed my options, but the list wasn’t empty. 

For example in this category, we have Meraki, which my work use and they have Meraki switch set at my house for my home workstation. Other brands include such as Aruba,  Ruckus and Juniper. Meraki and Juniper looked to have no subscription or license-free option so they were out.

On the other hand, both Ruckus and Aruba offers option to use enterprise gears without license or subscription. What’s great about these options are they use real enterprise hardware, but gives local access point based controller without recurring fee.

It is important to note that Aruba here is NOT same as Aruba Instant-On. They call Instant Mode. It’s a mode of operation for their enterprise Access Point. I really think they named these poorly because when I try to search about Instant-On I didn’t really catch Instant Mode and when I try to search Instant Mode operation reviews, I get bunch of Instant-on related sites. As a result, I initially thought Aruba did not offer enterprise product to run without subscription or license.

Basically, both Ruckus unleashed and Aruba Instant Mode are their simplest mode of control. The exact same devices can be upgraded to cloud based or dedicated device based controller for full-blown enterprise network system.  

Although I have zero intention for doing so, this felt really nice as I would be using the real enterprise products if I were to choose these.

Both of these systems looked promising for potentially offering overall better performance than UniFi WiFi systems given their track records, reputation, and most importantly the purpose of these devices were designed for.

So let’s take a closer look at two.

Price

Any time I try to search similarly spec’ed access points between these two brands, I find Aruba have better pricing. This is generally true on both Amazon and Ebay even on used ones. 

For instance, the cheapest I could find brand new Ruckus R750, which I believe practically the top-end model for Unleashed on Ebay from recently sold history was over $500 and even used were over $400. Whereas, the Aruba AP-535 model has new one on Bay for just $200. So that’s essentially half the price. 

Lower end WiFi 6 Access Point model from Ruckus is R350. The brand new R350 is $200 for recently sold Ebay. While AP-505 is listed $160 on Ebay. 

Winner: Aruba

Scalability

As mentioned earlier, part of my criteria is having an option to try out cutting edge technology. 

In this regard both vendors have WiFi 6E access points available to purchase, but Ruckus Unleashed firmware is not compatible with any of WiFi 6E models. So there is technically no Ruckus option for WiFi 6E.

However, this does not mean Aruba is superior in this category. Because Ruckus already have WiFi 7 AP listed on their website including actual specification sheet. If all planned, they should be releasing it by end of December.

Whereas, Aruba currently does not have any WiFi 7 access point. So unless Aruba release WiFi 7 Access Point before Ruckus, the order of who adopts the most cutting edge technology will change.

Basically, here I think as a home user, both should be fine from this point.

Performance

So far, it looks like Aruba could be the way to go, but what about the top most criteria for me, the stability and actual performance of access device?

This is indeed I spent most of my time researching. However, this maybe due to target market difference, getting actual performance related comparison data amongst enterprise gears were much harder than consumer gears.

PeerSpot

One source I found is peerspot.com where they look like compiling user reviews and use that to compare products.

According to this, Ruckus is listed as superior solution.

“Ruckus Wireless seems to be the superior solution, based on the parameters we compared. Our reviewers found Aruba Wireless to have stability issues and mentioned a poor ROI because of issues that needed fixing caused by the solution. Ruckus Wireless did better in these categories.”

https://www.peerspot.com/products/comparisons/aruba-wireless_vs_ruckus-wireless

Although I do not know the site’s credibility, they do mention price difference appropriately and of course giving Aruba its credit. Though this isn’t much hard to gather info, and doesn’t confirm credibility, at least, this does not disprove their credibility. 

On this comparison, they say users of Aruba like the GUI and scalability, but complaining the stability. In contrast,  users of Ruckus claims the performance being amazing and stability being great. Then complains about the security needing improvement. 

As a home user, I would never get the full GUI access to either of these products, so that’s irrelevant. Security could be major issue, but I think is is less relevant for home WiFi setting as we have full control of what’s connecting to our WiFi compared to public or enterprise setting. 

Certainly seeing my keyword of “stability” mentioned as weakness on Aruba, and strength on Ruckus makes me think Ruckus is better solution for me if this was trustworthy source.

Packet 6 Stress Test

Next, I came across a Youtube video “stress test” comparing Ruckus Wi-Fi 6 access points specifically R750 model against compatible models from other major enterprise brands including Aruba AP-535 as well as others such as Meraki MR46 and Mist AP43.  

The main conclusion of the test was only the Ruckus R750 AP was able to handle all traffic consistently with higher average throughput under the multiple concurrent device use.

This would have been the exact source I was looking for except…

This study was only sponsored, designed and conducted by Ruckus. The report was created by third party author who states he observed and confirmed the validity of the test.

I read actual test report, which is available free online. So let’s take a brief look at this.

Test setup

In brief, the test consisted 30 dell laptop simultaneously streaming 1080 HD video through local network via a single access point.

During the 3 minutes stress test period, 5 iPads simulating VoIP conversation, 5 iPad downloading data, and 20 MacBook Pro simulating varying size data download were introduced. All but 20 MacBook Pro were in the same Room as the Access point and MacBook Pro were in adjacent room. 

In order to reduce the bias, each AP used out-of-the-box default setting. 

Relevant measurements

During the stress test period, a few things were measured, which included # of stall-free streaming dell laptops and average throughput of downloading clients.

Results

Average throughput of Ruckus were 441 Mbps while Aruba came forth with 232 Mbps. Ruckus was completely stall-free with no laptop lost streaming based on the testing criteria while Aruba had almost half of the laptop stalled, yet it was the second best AP in this measurement.

Conclusion

Overall the test conclusion was despite all Access Points having similar specs, Ruckus R750 outperformed all other APs in every test.

Again, if this was a true third party designed and conducted test, this would have been close to the best proof confirming the Ruckus access point quality.

Study limitations?

Now the test could still be valid, but a few questions came to my mind, which weren’t addressed on the paper.

For example, although the test was done with out of the box setting for all APs, there is nothing here to prevent Ruckus from making the stress test condition identical to company’s normal internal device testing condition that are used to make default out-of-the box setting. If this were the case, the default out-of-the-box setting for the Ruckus AP is actually the same as optimized setting for testing condition.

Another point is the switch used in this test was  RUCKUS’s own. Since all clients were on the same VLAN, it might not have had much impact, but it would have been less confounding factor if they chose some well known third party switch or at least tested with couple different brand switches.

Third, the test was run for just 3 minutes. This felt a rather short time. Could the result change with longer stress? Perhaps Ruckus may perform well at beginning but may not be able to sustain its advantage for longer time?

I am certainly taking the result with a grain of salt, but it’s the only data driven comparison I could find from recent years between Aruba and Ruckus. 

My choice

Had I known Aruba Instant mode existence, the decision would have been much harder. 

But after pondering, I would have still chosen Ruckus because of my top priority was improve WiFi stability, and Ruckus seems to have the best chance based on my research. If I had chosen anything else, and if the performance gain is minimal when compared to UniFi, I would then have to keeping thinking about possibly of the better alternative for me. I won’t go in detail about Ruckus network’s performance comparing to UniFi here, because that’s for future write up. But just about a week of use, I am very impressed.

Update 12/26/2023

Soon after playing with Ruckus Unleashed, I found a good deal about Aruba Instant. So I ended up purchasing and started to test Aruba Instant as well. If you are interested in the specific comparisons amongst Aruba Instant, Ruckus Unleashed, and UniFi multi-access point setup, please see the dedicated series.