BNGSOFT Internet Acceleration
Accelerate Your Internet Experience
- Home
- Internet Acceleration
Popular objects served from your own storage instead of your transit link.
One subscriber serves another directly, without the traffic ever leaving your network.
Torrent traffic comes from many peers, so ordinary caches miss it. This does not.
Billed on the traffic it keeps off your upstream, at a fraction of transit price.
Two mechanisms, both transparent to the subscriber: hold popular content locally, and let subscribers serve each other instead of reaching across your transit link.
Two mechanisms that keep traffic off your transit link, and one honest boundary around what they can reach.
When one subscriber already holds a piece another is asking for, the transfer happens inside your network at local speed — with no extra hardware at either end and nothing installed on the CPE.
Popular objects are held on the node and served from there. Both plain and encrypted torrents are handled, alongside plain HTTP.
Content served locally arrives at local latency and local throughput, rather than across a transit link that is busiest exactly when your subscribers are.
Software updates, OS images, game downloads, static assets and P2P. TLS traffic — HTTPS, HTTP/2 and QUIC — passes through untouched, so the benefit tracks how much of your mix sits outside TLS.
Few systems save anything on P2P, because there is no single origin to cache. Routing those connections to local peers is what turns torrent traffic from an expensive line item into a cheap one.
It sits above the BNG in the path to your upstream. No client software, no proxy settings, no change to subscriber equipment, and nothing for support to explain.
Transparent caching and local peering, sitting above your BNG. It saves you transit on the traffic it can still see — and it is billed on exactly that, so it costs you nothing on the traffic it cannot help with.
It takes traffic off your transit link. Content served from the local cache, or from another subscriber inside your own network, never crosses the upstream — so you stop paying transit for the same object twice. It also gives you visibility of what your subscribers are actually pulling.
An L7 classifier recognises HTTP and P2P protocols and routes the request to the local cache, or to another subscriber that already holds the data, instead of out over the expensive link. It is transparent: nothing changes on the subscriber side and no client software is involved.
Because most caching products cannot help with it. Torrent traffic does not come from one server — it comes from many peers, so there is no single origin to cache. This system routes those connections to local peers or to the cache instead, which on a residential network is where the bulk of the saving comes from.
Above the NAS / BNG / BRAS, in line with the upstream. It is a separate product from the XDP BNG data plane and does not depend on it — it will sit in front of whatever access concentrator you already run.
It ships in two shapes: AIO, a single all-in-one node, and GRID, a scale-out set of nodes sharing one cache. Which you need follows from your busy-hour upstream rate and how much content you want to hold locally. Send us your peak throughput and cache-size target and we will size it properly rather than guess.
Plain HTTP objects, and both plain and encrypted torrents. It cannot touch TLS — HTTPS, HTTP/2 and HTTP/3 over QUIC are encrypted end to end and pass straight through untouched. That is a property of the protocols, not something we can engineer around, so the honest way to size the benefit is to measure how much of your traffic still sits outside TLS.
On traffic saved — the bytes the system served that would otherwise have crossed your transit. The price per saved gigabyte is a fraction of transit price, so it funds itself out of the saving rather than out of a capital budget. Ask for a quotation.
sales@bngsoft.com
Six things the subscriber edge has to do, and one forwarding path that does them — the same XDP data plane on the same commodity server, with no separate appliance per function. Each card says plainly where it stands.
PPPoE and IPoE termination with RADIUS AAA, CoA applied to the live forwarding plane, QinQ access and per-subscriber hierarchical shaping.
Port-block NAT44 in the forwarding path — no per-flow connection tracking, one log record per block, and idle reclaim that tightens as the pool fills.
Transparent caching and local peering for the traffic that is still cacheable — HTTP objects and P2P — billed on the upstream traffic it saves you.
White-label streaming platform: CMS and playout, apps for Android and Android TV, VOD and live channels, visible and forensic watermarking.
Transit forwarding with eBGP and FRR as the routing authority, on the same commodity server. Design complete and reviewed — not yet running in a network.
BCP 38 source validation, ACL chains and victim-and-vector DDoS containment — all in the same pass that forwards the packet, with no scrubbing hop.
There is no support tier between you and the engineers who built the forwarding path. When you report something in the data plane, it reaches the person who wrote that code path. More about how we work →
Acceleration is one of three things that decide whether a line feels fast. These cover all of them.
Send us a week of your traffic profile and your current transit commit. We will tell you honestly what share of it caching and local peering can take off the bill, and what it would cost to find out on your own network.
Contact us if you need a solution.