Educational Resource: This platform provides structured information about mobile data systems and connectivity — no account, recharge, or balance services are offered.

Connectivity Explained

What does it truly mean for a mobile device to be "connected"? Explore the technical mechanisms of signal acquisition, session maintenance, handover, and the conditions that disrupt or sustain a live data connection.

What Is Mobile Connectivity?

In technical terms, mobile connectivity is the operational state in which a device maintains an active EPS (Evolved Packet System) session with the network, has an allocated IP address, and possesses an active data bearer through which packets can flow to and from the public internet.

Connectivity is not a binary state — it exists on a spectrum. A device may be registered on the network (control plane active) but have no active data bearer (user plane inactive). It may have a data bearer active but be throttled to negligible speeds by policy enforcement. Or it may have full, unrestricted data access with a high-quality radio connection. Each of these represents a different point on the connectivity spectrum.

The word "connected" as users commonly understand it — being able to browse the web, use apps, and stream content at acceptable speeds — requires alignment of several simultaneous conditions: adequate radio signal quality, active EPS attachment, an operational data bearer, and an active data allocation (quota) in the charging system.

Radio Layer

Adequate SINR for data transmission via OFDMA radio channels

Authentication Layer

Valid EPS session with active SIM authentication and bearer allocation

Policy Layer

Active data quota in OCS with PCEF gate status set to OPEN

Signal Acquisition and Cell Selection

Before any data can flow, a mobile device must locate and register with a suitable cell. This process — called cell selection — involves the device's baseband modem scanning available radio frequencies, measuring signal quality from multiple base stations, and selecting the best candidate according to 3GPP-specified criteria.

The primary metric used in cell selection is Reference Signal Received Power (RSRP) — a measurement of the power level of the LTE reference signals transmitted by the base station. A secondary metric, Reference Signal Received Quality (RSRQ), accounts for interference from neighbouring cells. Together, these metrics help the device and network determine the optimal serving cell at any given moment.

RSRP Range (dBm)Signal QualityExpected Data Experience
> −80 dBmExcellentMaximum throughput, highest modulation (256-QAM)
−80 to −90 dBmGoodHigh throughput, minimal packet loss
−90 to −100 dBmFairModerate throughput, occasional retransmissions
−100 to −110 dBmPoorLow throughput, significant retransmissions, high latency
< −110 dBmVery Poor / No ServiceConnection failures, data session drops

Once a suitable cell is selected, the device performs Random Access (RACH procedure) to request uplink resources from the base station's MAC scheduler. A successful RACH completes the radio connection and allows the subsequent Attach procedure to commence.

Handover Mechanisms

One of the most technically impressive aspects of mobile connectivity is the handover — the seamless transfer of a device's connection from one base station to another as the device moves through coverage areas. Handovers happen automatically, typically without any interruption the user would notice, and may occur hundreds of times during a single journey.

X2 Handover (Direct eNodeB Transfer)

The most common handover type in LTE is the X2 handover, executed directly between two eNodeBs that are connected via the X2 interface. The serving eNodeB monitors the device's signal measurements (reported by the UE via Measurement Reports), and when a neighbouring cell consistently exceeds the configured A3 event threshold (neighbouring cell RSRP is stronger than serving cell by defined margin), the serving eNodeB initiates the handover.

01

Measurement Report Triggered

The device sends a Measurement Report to the serving eNodeB, indicating that a neighbouring cell (target) exceeds the A3 event trigger threshold (e.g., target RSRP is 3dB stronger than serving cell).

02

Handover Request to Target eNodeB

The serving eNodeB sends a Handover Request to the target eNodeB via the X2 interface, including the device's context (bearer configuration, security keys, QoS parameters).

03

Target Cell Preparation

The target eNodeB allocates resources for the incoming device and responds with a Handover Request Acknowledge, including a transparent container with the new cell's radio configuration.

04

Handover Command & Random Access

The serving eNodeB sends a Handover Command to the device. The device performs random access on the target cell, confirms successful handover, and begins communicating with the new base station.

Handover Speed

A well-executed X2 handover typically completes in 50–80 milliseconds — imperceptible to the user. During this window, the serving eNodeB buffers downlink packets and forwards them to the target eNodeB to minimise data loss. The device's IP address, active bearers, and OCS session all remain unchanged throughout.

Session Maintenance: Keeping the Connection Alive

Maintaining an active mobile data session requires continuous background activity between the device, the network, and the charging systems. Even when no application is actively transmitting data, several mechanisms operate to keep the session alive and ready for instant data transfer.

RRC (Radio Resource Control) States

The LTE radio interface operates in two primary states: RRC_CONNECTED (active radio bearer, lowest latency for data) and RRC_IDLE (no active radio bearer, device monitors paging channel). Transitioning from IDLE to CONNECTED — called RRC Connection Re-establishment — takes approximately 100–200ms.

To conserve battery power, devices automatically transition to RRC_IDLE when no data has been transmitted for a configurable inactivity timer period (typically 10–30 seconds). This does not terminate the EPS session — the IP address and bearers are preserved at the core network level.

DRX: Discontinuous Reception

Connected-mode DRX (C-DRX) allows a device in RRC_CONNECTED state to enter periodic sleep cycles between data transmissions, reducing power consumption without fully releasing the radio connection. The device "wakes up" at configured intervals to check for incoming data.

DRX cycles are negotiated between the device and the network and are invisible to applications. The trade-off is a slight increase in uplink latency (the device may be asleep when data arrives) versus significant battery life savings — critical for always-on applications like push notifications.

Keep-Alive and IP Session Persistence

At the IP layer, many applications and operating systems send periodic keep-alive packets — small TCP or UDP datagrams — to prevent Network Address Translation (NAT) state tables from expiring. This is particularly important on mobile networks where CGNAT is used: if no traffic is observed for a NAT entry's timeout period (typically 30–300 seconds for UDP, 30–120 minutes for TCP), the NAT mapping is removed and the application's connection effectively breaks.

Causes of Connectivity Disruption

Connectivity disruptions on mobile networks arise from a variety of causes — spanning the radio layer, the network infrastructure layer, and the policy/charging layer. Understanding these causes helps contextualise why data access may intermittently fail or degrade.

When a device moves beyond the coverage boundary of all available cells, or when building materials, terrain, or interference cause RSRP to fall below the minimum threshold (~−120 dBm), the device loses radio connectivity. The EPS session is suspended but not immediately terminated — core network timers (typically 30–60 seconds) allow the device to reconnect without a full re-attach if signal is re-acquired quickly. Persistent out-of-coverage causes the MME to release the session, requiring a full re-attach when coverage is regained.

When the OCS quota counter reaches zero, the PCEF applies the operator's exhaustion policy — typically a hard block (gate CLOSED) or throttle. The device remains attached to the network and its radio connection is unaffected — only the data plane policy changes. This is the connectivity disruption caused by data balance exhaustion. It is resolved by a recharge event that restores the OCS quota and triggers a PCEF policy update to re-open the data gate. See our recharge concept explanation for the full technical detail.

In heavily loaded cells, the MAC scheduler has a finite number of resource blocks to distribute among active devices. Under congestion, each device receives fewer resource blocks per scheduling interval, reducing effective throughput. Operators manage congestion through capacity planning, carrier aggregation, and dynamic spectrum sharing — but in peak periods (stadium events, rush hours), users may experience significantly degraded speeds even with full signal strength.

While most handovers complete seamlessly, a small percentage fail — typically due to rapid signal changes (fast vehicle movement), misconfigured handover parameters, or radio resource contention on the target cell. A failed handover causes a Radio Link Failure (RLF), which triggers an RRC Re-establishment procedure. If re-establishment fails, the device must perform a full re-attach, briefly losing connectivity. The entire recovery process typically takes 1–5 seconds.

Beyond data quota exhaustion, a subscriber's overall plan may expire (e.g., a 30-day prepaid plan reaches its expiry date). In this case, the operator's BSS updates the subscriber's service profile in the HSS and PCRF to remove data access permissions entirely — not just quota-gate it. The PCRF pushes a new policy to the PCEF blocking all data services. Restoring access requires a plan renewal transaction in the operator's billing system, which follows the same BSS–OCS–PCEF update path as a standard recharge.

Complete Connectivity State Model

The following diagram represents the complete connectivity state model for a mobile device, combining radio states, EPS session states, and data policy states into a unified view.

Mobile Device — Connectivity State Model
State: FULLY CONNECTED
RRC_CONNECTED + Active Bearer + OCS Quota Available
Optimal
Full data access at negotiated radio speed. All network layers functional. User can browse, stream, and use apps normally.
State: IDLE — SESSION PRESERVED
RRC_IDLE + EPS Session Active + OCS Quota Available
Standby
No active radio bearer. Device monitors paging. Reconnects to RRC_CONNECTED within ~200ms when data is required. IP address preserved.
State: THROTTLED
RRC_CONNECTED + Active Bearer + FUP/Low Quota Policy Applied
Degraded
Data flows but at reduced MBR (e.g., 128kbps–1Mbps). Policy enforced by PCEF via updated QoS profile. Basic browsing possible; streaming typically fails.
State: DATA SUSPENDED
Attached + PCEF Gate CLOSED (Quota Exhausted)
Suspended
Device is attached; SIM authenticated; voice/SMS functional. Only data plane blocked. Resolved by recharge/top-up triggering OCS quota restoration and PCEF gate re-open.
State: NO COVERAGE
No Radio Signal — EPS Session Pending Release
Offline
Device out of range of all cells. MME starts T3412 timer (default 54 minutes). Session released if device does not re-attach within timer period. Full re-attach required on coverage regain.

Continue with Our Blog Articles

Read in-depth articles on how mobile networks function in daily operation and explore digital connectivity concepts in practical terms.