WAN Concepts and Encapsulation
Key Takeaways
- A WAN connects geographically separated sites across a service-provider network; a LAN is a single administrative domain with short-distance Ethernet switching.
- Common WAN access methods at HCIA level include leased lines, broadband (DSL/cable/fiber with PPPoE), and cellular (3G/4G/5G) uplink for branches and backup.
- WAN links need Layer 2 encapsulation (HDLC, PPP, Ethernet/PPPoE) so frames can cross the provider circuit and carry Layer 3 packets.
- PPP uses ordered phases: LCP (link establishment and options), optional authentication, then NCP (for example IPCP) to assign or negotiate network-layer parameters.
- MPLS and Segment Routing (SR) are provider core forwarding technologies you should recognize at concept level—not design in depth on HCIA-Datacom.
WAN Concepts and Encapsulation
Quick Answer: A WAN joins distant sites across a service-provider network; a LAN is local, high-speed Ethernet under one admin. WAN circuits need encapsulation (HDLC, PPP, Ethernet/PPPoE) so IP packets can ride the link. PPP brings the link up with LCP, may authenticate, then runs NCP (for example IPCP) for network parameters. MPLS and Segment Routing are provider-core concepts at awareness level only on HCIA.
Campus switching and IP routing matter only if remote sites can reach each other. HCIA-Datacom expects you to place WAN technology correctly in a design, choose encapsulation ideas for serial or broadband access, and avoid confusing enterprise edge methods with provider-core MPLS. This section builds those foundations before PAP/CHAP and PPPoE configuration sections go deeper.
WAN versus LAN
| Attribute | LAN | WAN |
|---|---|---|
| Geographic scope | Building, campus, or short campus interconnect | City, country, or global site-to-site |
| Typical media | Copper/fiber Ethernet, Wi-Fi | Leased circuits, broadband last-mile, cellular, provider Ethernet |
| Who owns the path | Enterprise (switches, cabling) | Mix of enterprise CPE and service-provider infrastructure |
| Bandwidth cost | Relatively high capacity, lower per-bit cost | Often lower capacity and higher cost per bit |
| Addressing focus | Private enterprise subnets, VLANs | Public/private peering, edge addressing, often NAT at the edge |
| Primary protocols (exam view) | Ethernet, STP/LAG, VLAN, OSPF inside site | PPP/PPPoE, static/default toward ISP, optional VPN |
Exam language trap: “WAN” does not mean “any slow link.” A 10 G campus backbone between two buildings on the same estate is still a LAN design problem. A branch office connecting through an ISP over hundreds of kilometers is a WAN problem even if the access link is fiber broadband.
From an enterprise perspective, the customer edge (CE) device—often a Huawei AR-series router on exam diagrams—terminates the WAN access technology and runs routing (static default, OSPF toward the LAN, or simple dual-homing). The provider side may present a serial-like circuit, Ethernet handoff, or PPPoE session endpoint.
Why encapsulation is required
Layer 3 (IP) needs a Layer 2 frame format to leave an interface. On Ethernet LANs that format is IEEE 802.3/Ethernet II. On classic serial WAN links, or on broadband sessions that reuse PPP, the device must speak a WAN data-link protocol so both ends agree on:
- Frame delimiters and error detection (FCS/CRC concepts)
- How keepalives prove the link is up
- Optional authentication before carrying user traffic
- How IPv4 (and sometimes IPv6) is mapped onto the link
Without matching encapsulation on both ends of a point-to-point circuit, the physical line may show “up” while Layer 2 stays down and no IP adjacency forms.
Common WAN access methods (HCIA overview)
| Access type | Typical use | What the enterprise router sees | HCIA emphasis |
|---|---|---|---|
| Leased line / dedicated circuit | Stable site-to-site or site-to-provider | Serial or Ethernet private circuit; often PPP or HDLC style framing historically | Encapsulation and point-to-point IP |
| Broadband (DSL, cable, FTTH) | Home, SME, many branch Internet edges | Ethernet handoff; frequently PPPoE client toward ISP | PPPoE discovery + PPP session |
| Cellular (3G/4G/5G) | Primary remote access or backup uplink | Cellular modem/interface; IP over cellular data session | Concept: wireless WAN backup, not RF engineering |
| Provider Ethernet / Metro Ethernet | Campus-to-DC or multi-site L2/L3 handoff | Ethernet port; may still run PPP only if ISP requires it | Treat as Ethernet WAN edge unless stem says PPP |
Leased lines historically used serial interfaces with clocking and PPP/HDLC. Modern campuses often receive an Ethernet handoff that still behaves as a point-to-point WAN for routing purposes: one next hop toward the ISP or peer site.
Broadband access is extremely common for small and midsize branches. The ISP frequently requires a username/password session. That is why PPPoE appears on H12-811: Ethernet locally, PPP for the session and authentication toward the provider.
Cellular appears as an overview option: a router with a cellular interface can install a default route when the primary fiber or leased path fails. Exam items test the role (backup WAN, remote site) more than modem AT commands.
PPP as the workhorse WAN encapsulation
Point-to-Point Protocol (PPP) is a standard data-link protocol for point-to-point links. It is media-flexible (serial, Ethernet via PPPoE, and other tunnels in broader networking) and supports authentication, multilink concepts (beyond deep HCIA labs), and network-layer negotiation.
PPP phases you must sequence correctly
Think of PPP as a state machine, not a single “on” switch:
| Phase | Name | What happens |
|---|---|---|
| 1 | LCP (Link Control Protocol) | Peers establish the link, negotiate options (MRU, authentication protocol, magic number for loop detection, compression flags, etc.), and exchange keepalives |
| 2 | Authentication (optional but common) | After LCP agrees that authentication is required, peers run PAP or CHAP (next section) |
| 3 | NCP (Network Control Protocol) | Per-protocol network setup; for IPv4 this is IPCP, which can negotiate IP addresses and related options |
Exam order trap: Authentication is not the first step. LCP succeeds first, then authentication (if configured), then NCP/IPCP. If authentication fails, NCP should not complete and IP will not run on that PPP session.
LCP failure symptoms in lab language: interface protocol stays down, keepalives fail, or peers disagree on required options. NCP failure after successful auth often looks like “link is up but no IP” or missing negotiated address from the ISP.
HDLC comparison (brief)
HDLC (High-Level Data Link Control) family encapsulations also carry frames on serial point-to-point links. Vendor HDLC variants (including historically Cisco-style HDLC on serial) are simple and do not provide the same standardized multi-protocol authentication and negotiation model as PPP.
| Topic | PPP | HDLC (conceptual) |
|---|---|---|
| Standardization for multi-vendor auth | Strong (PAP/CHAP defined with PPP) | Weaker / vendor-specific variants |
| Option negotiation | Rich LCP/NCP model | Minimal compared with PPP |
| Typical exam preference when auth is required | PPP | Not the answer when username/password auth is required |
| Complexity | Higher (phases, auth, NCP) | Lower |
If a question requires PAP or CHAP, the encapsulation context is PPP (or PPPoE, which carries PPP). HDLC is not the path for those authentication methods.
Enterprise edge versus provider core: MPLS and Segment Routing
HCIA-Datacom may mention MPLS (Multiprotocol Label Switching) and Segment Routing (SR) as modern WAN/backbone ideas. Stay at basic concepts:
- MPLS forwards packets using short labels rather than performing a full IP longest-match lookup at every hop in the provider core. Providers build label-switched paths; enterprises often only see an IP handoff or a VPN service edge.
- Segment Routing encodes the path as an ordered list of segments (for example node or adjacency SIDs), simplifying traffic engineering compared with some classical MPLS control-plane approaches.
| Idea | What HCIA needs you to know |
|---|---|
| Who runs MPLS/SR | Usually the service provider core, not the campus access switch |
| Enterprise touchpoint | CE router peers with provider PE using IP/Ethernet/PPP as designed; VPN service may hide the core |
| Exam trap | Do not “configure MPLS TE” as the answer to a branch PPPoE Internet question |
| Relation to this chapter | Encapsulation and access methods at the edge; MPLS/SR as awareness of how large WANs scale |
You are not expected to design RSVP-TE, LDP meshes, or SR policy stacks at HCIA depth. Recognize the vocabulary so distractors do not steal marks from PPP/PPPoE questions.
Putting it together in a simple branch design
A typical small-branch picture for this chapter:
- LAN users sit behind a Huawei router with VLANIF or routed interfaces and private addressing.
- The WAN interface uses PPP on a leased-style circuit or PPPoE on an Ethernet broadband handoff.
- After the session is up, the router installs a default route toward the ISP (static or learned via IPCP/other means depending on design).
- Optional cellular interface provides a floating default with worse preference for backup.
- Inside the campus, OSPF or static routing continues as studied in earlier chapters; WAN is just another next-hop domain with different encapsulation.
Exam traps checklist
- LAN vs WAN is about scope and provider path, not only speed.
- Encapsulation mismatch kills Layer 2 even when the cable is fine.
- PPP order is LCP → authentication → NCP, never auth-first.
- Need username/password on a WAN link → think PPP/PPPoE, not bare HDLC.
- Broadband home/SME Internet → PPPoE is a frequent correct access method.
- MPLS/SR = provider core awareness, not the substitute for edge PPP knowledge.
Practice goals before the next section
- Explain in one sentence how a branch WAN differs from a campus LAN backbone.
- List three access methods and the encapsulation each commonly implies.
- Order LCP, CHAP, and IPCP for a successful IPv4 PPP session.
- State why an exam item that requires authentication points to PPP rather than HDLC.
Next, you will compare PAP and CHAP in detail—the authentication phase that sits between LCP and NCP on real ISP and enterprise PPP links.
Which statement correctly contrasts a LAN with a WAN for HCIA-Datacom design language?
What is the correct order of PPP phases when authentication is configured for an IPv4 session?
An ISP requires a username and password on a serial point-to-point access circuit. Which encapsulation family is appropriate?
How should Multiprotocol Label Switching (MPLS) and Segment Routing (SR) be treated on the HCIA-Datacom exam in the WAN chapter?